Datasets:

G-reen
/

fastdetector-train-stat-train

original stringlengths 226 30.2k	prompt dict	topic stringclasses 23 values	format stringclasses 21 values	response_0 stringlengths 11 59.1k	response_1 stringlengths 0 186k	generator_model stringclasses 1 value	generation_params stringclasses 1 value	final_response stringlengths 51 159k	jaccard_1 float64 0 1	jaccard_2 float64 0.02 1	levenshtein float64 1 158k	cosdist float64 -0.01 1	original_editlens_bucket_roberta_large int64 0 3	original_editlens_score_roberta_large float64 0.01 1	final_response_editlens_bucket_roberta_large int64 0 3	final_response_editlens_score_roberta_large float64 0.01 1	original_editlens_score_llama_3_2_3b float64 0 0.78	final_response_editlens_bucket_llama_3_2_3b int64 0 3	final_response_editlens_score_llama_3_2_3b float64 0 1	original_editlens_score_giga_roberta float64 0 1	final_response_editlens_bucket_giga_roberta int64 0 3	final_response_editlens_score_giga_roberta float64 0 1	original_editlens_score_giga_llama_3_2_3b float64 0 0.79	final_response_editlens_bucket_giga_llama_3_2_3b int64 0 3	final_response_editlens_score_giga_llama_3_2_3b float64 0 1
Jamie's Italian will be hosting a special Jamie Cooks Italy Feast event on Wednesday the 27th of March 2019. The evening will include a delicious menu developed by Jamie from his trip across Italy. Dishes include tuna carpaccio, bucatini all'amatricana, fiorentina steak and amalfi lemon tart. The tuna carpaccio has bee...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nClarify this text by improving the phrasing while preserving its original intent.\nOutput the full new text with no extra statements or commentations.\nBegin directly with the text itself. Do not add a title, a heading, or a label naming what you have written...	Food & Dining	Product Page	Jamie’s Italian will be hosting a special Jamie Cooks Italy Feast on Wednesday, 27 March 2019. The evening features a menu inspired by Jamie’s journey across Italy, showcasing regional favorites. The menu includes tuna carpaccio, bucatini all’amatriciana, fiorentina steak, and amalfi lemon tart. The tuna carpaccio is p...		ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	Jamie’s Italian will be hosting a special Jamie Cooks Italy Feast on Wednesday, 27 March 2019. The evening features a menu inspired by Jamie’s journey across Italy, showcasing regional favorites. The menu includes tuna carpaccio, bucatini all’amatriciana, fiorentina steak, and amalfi lemon tart. The tuna carpaccio is p...	0.673469	0.821596	344	0.130649	0	0.01389	1	0.294343	0.019795	1	0.557815	0.015475	2	0.629033	0.020436	3	0.967778
From undiscovered designers to your most loved labels, Studio 12.20 creates an unrivaled collection that encompasses clothing, footwear, and accessories, all at amazing prices while never sacrificing quality! Studio 12.20 San Diego first launched in the heart of eclectic Hillcrest at 1220 Cleveland Avenue in 1996. Stud...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nEdit this text by inserting exactly one to three sentences (in appropriate locations), you may only make minimal modifications to transitions in adjacent sentences for better flow.\nOutput the full new text with no extra statements or commentations.\nBegin di...	Fashion & Beauty	About (Org.)	From undiscovered designers to your most loved labels, Studio 12.20 creates an unrivaled collection that encompasses clothing, footwear, and accessories, all at amazing prices while never sacrificing quality! Studio 12.20 San Diego first launched in the heart of eclectic Hillcrest at 1220 Cleveland Avenue in 1996. Stud...	From undiscovered designers to your most loved labels, Studio 12.20 creates an unrivaled collection that encompasses clothing, footwear, and accessories, all at amazing prices while never sacrificing quality! Studio 12.20 San Diego first launched in the heart of eclectic Hillcrest at 1220 Cleveland Avenue in 1996. Stud...	ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	From undiscovered designers to your most loved labels, Studio 12.20 creates an unrivaled collection that encompasses clothing, footwear, and accessories, all at amazing prices while never sacrificing quality! Studio 12.20 San Diego first launched in the heart of eclectic Hillcrest at 1220 Cleveland Avenue in 1996. Stud...	0.054167	0.078488	18	0.012237	0	0.10017	0	0.108482	0.010792	0	0.015216	0.001825	0	0.002152	0.002611	0	0.002516
How can you progress from a film idea to a film project, from the dream of a film to a film-to-be, how can you reach your objective: These are crucial questions for a documentary filmmaker. In-depth working on the scriptwriting, designed as a tool for the film development, should enable to answer or at least clarify mo...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nEdit this text to eliminate redundant and filler words.\nOutput the full new text with no extra statements or commentations.\nBegin directly with the text itself. Do not add a title, a heading, or a label naming what you have written.", "{{RESP_0}}\nExpan...	Education & Jobs	About (Org.)	Progressing from a film idea to a film project—from a dream to a film-to-be—is crucial for documentary filmmakers. In-depth scriptwriting, used as a development tool, can answer or clarify most such questions. The Lussas residency, founded in 1997, welcomes documentary scriptwriters, experienced or not, offering a shar...	Progressing from a film idea to a film project—from a dream to a film-to-be—is crucial for documentary filmmakers. In-depth scriptwriting, used as a development tool, can answer or clarify most such questions. The Lussas residency, founded in 1997, welcomes documentary scriptwriters, experienced or not, offering a shar...	ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	Progressing from a film idea to a film project—from a dream to a film-to-be—is crucial for documentary filmmakers. In-depth scriptwriting, used as a development tool, can answer or clarify most such questions. The Lussas residency, founded in 1997, welcomes documentary scriptwriters, experienced or not, offering a shar...	0.863924	0.939457	2,074	0.145866	0	0.037401	3	0.999333	0.031091	3	0.888735	0.006404	3	0.999983	0.005183	3	0.999116
Conspiracy, no. Cover-up, yes maybe. I'd sure call it a cover-up if after a company does something that could potentially cause big problems for others, they subsequently try to keep it quiet and/or has its account reps make up stories while they scramble to figure out a solution and damage control. One definition of a...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nChange the tone of this text to be more conversational and approachable.\nOutput the full new text with no extra statements or commentations.\nBegin directly with the text itself. Do not add a title, a heading, or a label naming what you have written.\nThe fi...	Finance & Business	Comment Section	So, let’s talk about this—call it a cover-up if you want, I’d probably agree. When a company does something that might cause real problems for others, and then they start downplaying it, have their reps give vague answers, or try to fix things quietly without being transparent? Yeah, that smells like a cover-up to me. ...		ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	So, let’s talk about this—call it a cover-up if you want, I’d probably agree. When a company does something that might cause real problems for others, and then they start downplaying it, have their reps give vague answers, or try to fix things quietly without being transparent? Yeah, that smells like a cover-up to me. ...	0.805907	0.95	1,863	0.146197	0	0.015395	2	0.627836	0.027899	3	0.850554	0.002516	3	0.880454	0.00251	3	0.814586
As Mike stood in the pub doorway the smoke from inside billowed out into the bitter wintry air. It swirled in the momentary light that streamed out until the door swung shut. He toppled down the step onto the cobbled street and the houses swayed like they were part of a fun house mirror illusion. After only six stagger...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nTranslate the given text to French.\nDo not output anything besides what you were requested to write, and do not output any extra commentary.", "{{RESP_0}}\nTranslate this text to English. Output only the English translation with no extra commentary.\nBeg...	Literature	Creative Writing	Comme Mike se tenait au seuil du pub, la fumée de l'intérieur s'échappe dans l'air froid et hivernal. Elle tourbillonnait dans la lumière passagère qui s'échappait jusqu'à ce que la porte se ferme. Il tomba sur l'escalier en pierre, les maisons semblant pencher comme dans une illusion de maison folle. Après seulement s...	As Mike stood at the threshold of the pub, the smoke from inside drifted into the cold, winter air. It swirled in the fleeting light until the door closed. He stumbled down the stone staircase, the houses seeming to tilt like in a madhouse illusion. After only six disoriented steps, he collapsed, vomiting onto the ston...	ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	As Mike stood at the threshold of the pub, the smoke from inside drifted into the cold, winter air. It swirled in the fleeting light until the door closed. He stumbled down the stone staircase, the houses seeming to tilt like in a madhouse illusion. After only six disoriented steps, he collapsed, vomiting onto the ston...	0.5594	0.709988	1,805	0.056813	0	0.018558	1	0.242717	0.046976	0	0.25682	0.010432	1	0.301277	0.097232	1	0.25206
Ekiti State High court sitting in Ado-Ekiti on Tuesday, September 29, sentenced 29-year-old Gbenga Aluko, to death by hanging for involving in cultism and killings of three residents in the state. Delivering his judgment, Justice Abiodun Adesodun said the prosecutor was able to prove the case beyond reasonable doubt ag...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nTake inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should use varied sentence lengths, and remove all common AI buzzwords such as 'tapestry', 'delve', or 'testament', to e...	Crime & Law	News Article	Court in Oyo State sitting in Ibadan on Thursday, October 10, sentenced 34-year-old Tolu Alaba to 15 years imprisonment for his role in a series of armed robberies targeting commercial buses along the Ibadan–Oyo expressway between June and August 2023. The presiding judge, Magistrate Kehinde Oladipo, said the prosecut...		ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	Court in Oyo State sitting in Ibadan on Thursday, October 10, sentenced 34-year-old Tolu Alaba to 15 years imprisonment for his role in a series of armed robberies targeting commercial buses along the Ibadan–Oyo expressway between June and August 2023. The presiding judge, Magistrate Kehinde Oladipo, said the prosecuti...	0.890052	0.974662	1,588	0.535523	0	0.008669	3	0.999498	0.009216	3	0.99927	0.002836	3	0.999984	0.001224	3	0.999996
I met Kölbel and Hilliger at the Asian restaurant beside the theatre. They dress similar colors, sitting right across me. They are both German and the founder of Berlin Circus Festival; Josa Kölbel studied in Circus art and performance arts in France as a trapeze catcher; John Hilliger’s background is event management....	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nAdapt this text for social media sharing.\nOutput the full new text with no extra statements or commentations.\nBegin directly with the text itself. Do not add a title, a heading, or a label naming what you have written.\nThe final text must have the format <...	Art & Design	Personal Blog	Last week, I met Kölbel and Hilliger at a small Asian restaurant just beside the theatre. They were sitting across from me, dressed in similar colors—calm, thoughtful, and clearly in sync. Both are German, founders of the Berlin Circus Festival, and have known each other since school. Josa Kölbel studied circus arts an...		ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	Last week, I met Kölbel and Hilliger at a small Asian restaurant just beside the theatre. They were sitting across from me, dressed in similar colors—calm, thoughtful, and clearly in sync. Both are German, founders of the Berlin Circus Festival, and have known each other since school. Josa Kölbel studied circus arts an...	0.733553	0.935663	9,269	0.067468	0	0.023502	1	0.556305	0.082727	3	0.765773	0.004717	2	0.596648	0.007637	1	0.471932
I have lived here my entire life and have developed a passion for what Alaska has to offer. I have spent my life fishing Alaska's rivers, biking, hiking the beautiful Alaskan trails and have become familiar with Alaska's real estate. I have dedicated myself to learning the difference in value for each property. I provi...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nWrite a new piece of text that uses this text as a citation or reference.\nOutput the full new text with no extra statements or commentations.\nBegin directly with the text itself. Do not add a title, a heading, or a label naming what you have written." ], ...	Home & Hobbies	About (Pers.)	Having grown up in Alaska and immersed in its natural beauty, I bring a deep personal connection to the real estate market, especially in areas popular for outdoor activities like fishing, biking, and hiking. Drawing from firsthand experience navigating the state’s unique landscapes and property values, I approach each...		ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	Having grown up in Alaska and immersed in its natural beauty, I bring a deep personal connection to the real estate market, especially in areas popular for outdoor activities like fishing, biking, and hiking. Drawing from firsthand experience navigating the state’s unique landscapes and property values, I approach each...	0.870466	0.986622	748	0.248258	0	0.018861	3	0.999506	0.004953	3	0.999816	0.008473	3	0.999895	0.003771	3	0.998173
If two figures have the same size and shape, then they are congruent. The term congruent is often used to describe figures like this. In this tutorial, take a look at the term congruent! If you need to find the distance between to points on the coordinate, you'll probably use the distance formula to get your answer. Th...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nWrite an alternate version of this text, it should follow the style and format of the original text.\nOutput the full new text with no extra statements or commentations.\nBegin directly with the text itself. Do not add a title, a heading, or a label naming wh...	Science & Tech.	Tutorial	If two figures have the same size and shape, then they are congruent. The term congruent is often used to describe figures like this. In this tutorial, take a look at the term congruent! If you need to find the distance between two points on the coordinate plane, you'll probably use the distance formula to get your a...		ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	If two figures have the same size and shape, then they are congruent. The term congruent is often used to describe figures like this. In this tutorial, take a look at the term congruent! If you need to find the distance between two points on the coordinate plane, you'll probably use the distance formula to get your ans...	0.03	0.075342	12	0.006118	1	0.371485	1	0.357568	0.057959	0	0.066065	0.068672	0	0.06804	0.043196	0	0.046902
for sure you will never be able to reach the quality of PhotoShop - they have their own algorithms, and I have my doubts they will release them as product. So if you need PhotoShop quality you need to see if you can automate the use in some manner. But there is another excellent image manipulation tool from the Unix/Li...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nRestructure this text to ensure a clear introduction, body, and compelling conclusion.\nOutput the full new text with no extra statements or commentations.\nBegin directly with the text itself. Do not add a title, a heading, or a label naming what you have wr...	Software	Comment Section	Software image manipulation and resizing solutions vary in quality, accessibility, and ease of integration, especially when working within environments like Visual Objects. Achieving the same level of quality as Adobe Photoshop is highly unlikely with open-source libraries, as Photoshop employs proprietary, highly opti...		ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	Software image manipulation and resizing solutions vary in quality, accessibility, and ease of integration, especially when working within environments like Visual Objects. Achieving the same level of quality as Adobe Photoshop is highly unlikely with open-source libraries, as Photoshop employs proprietary, highly opti...	0.865415	0.989139	3,601	0.175285	0	0.020873	3	0.889936	0.019237	3	0.807578	0.002139	3	0.960847	0.002249	3	0.9007
In a society of things, social responsibility requires a recognition of the influence of commodities upon our most foundational spiritual experiences. Nickelodeon's animated series, Rocko's Modern Life, puts it simply. Cartoons about capitalism are as old as cartoons themselves, but few have taken its absurdities as fa...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nWrite a new piece of text that is on a completely different topic, but reuses verbatim as many phrases and sentences from this text as possible.\nOutput the full new text with no extra statements or commentations.\nBegin directly with the text itself. Do not ...	Entertainment	Nonfiction Writing	In a society of entertainment, social responsibility requires a recognition of the influence of media on our most foundational cultural experiences. A popular streaming platform’s animated series, The Great Show, puts it simply. Cartoons about technology are as old as cartoons themselves, but few have taken its absurd...		ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	In a society of entertainment, social responsibility requires a recognition of the influence of media on our most foundational cultural experiences. A popular streaming platform’s animated series, The Great Show, puts it simply. Cartoons about technology are as old as cartoons themselves, but few have taken its absurdi...	0.151467	0.292854	1,817	0.368526	0	0.225026	1	0.262545	0.062407	0	0.052081	0.007669	0	0.061497	0.002914	0	0.014995
Parents and guardians of middle school and high school students, Hello. Hi. How are you doing? You good? Everything will be ok. We promise. But seriously I hope this blog finds you well! As you know, our Friday nights and Sunday nights have been canceled until further notice. For the time being, everything will be onli...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nModify this text to be as long as possible. At least half of the original text must be repeated verbatim, and at most three quarters of it; the remainder must be new prose.\nOutput the full new text with no extra statements or commentations.\nBegin directly w...	Religion	News (Org.)	Parents and guardians of middle school and high school students, Hello. Hi. How are you doing? You good? Everything will be ok. We promise. But seriously I hope this blog finds you well! As you know, our Friday nights and Sunday nights have been canceled until further notice. For the time being, everything will be onli...		ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	Parents and guardians of middle school and high school students, Hello. Hi. How are you doing? You good? Everything will be ok. We promise. But seriously I hope this blog finds you well! As you know, our Friday nights and Sunday nights have been canceled until further notice. For the time being, everything will be onli...	0.770335	0.821483	3,746	0.115171	0	0.027248	1	0.36669	0.011526	1	0.420814	0.006697	3	0.999953	0.004258	3	0.999993
No, it’s not a prequel to last year’s vertiginous Sundance prizewinner. In his sophomore feature, Ben Younger (Boiler Room) attempts to reinvent himself as a young Woody Allen, reversing the Woodman’s age and gender dynamics but yielding mortifying moments all the same. Two weeks after getting divorced, 37-year-old Raf...	{ "chat_turns": [ "<document>\n{{DOC}}\n</document>\n\nWrite another piece of text that resembles the style and language of this text.\nOutput the full new text with no extra statements or commentations.\nBegin directly with the text itself. Do not add a title, a heading, or a label naming what you have written.\...	Entertainment	Nonfiction Writing	In his follow-up to a breakout indie hit, the director tries to channel the aching sincerity of early Craig Gillespie, but ends up circling the same emotional cul-de-sac with less nuance and more self-awareness of its own clichés. The film centers on a 31-year-old publicist, played by Zooey Deschanel in a role that see...		ibm-granite/granite-4.2-30b-nvfp4	{"temperature": 1.0, "top_p": 0.95, "extra_body": {"chat_template_kwargs": {"enable_thinking": false}}}	In his follow-up to a breakout indie hit, the director tries to channel the aching sincerity of early Craig Gillespie, but ends up circling the same emotional cul-de-sac with less nuance and more self-awareness of its own clichés. The film centers on a 31-year-old publicist, played by Zooey Deschanel in a role that see...	0.910165	0.993559	1,618	0.524177	0	0.185534	3	0.999119	0.060488	3	0.999971	0.005671	3	0.999984	0.005437	3	0.999997

End of preview. Expand in Data Studio

Auto-generated FastDetector dataset

Fastdetector Train Stat Train

Best detector

Giga EditLens Llama-3.2-3B Score

0.7121 TPR @ 1% FPR

Hardest prompt subset

rewrite

0.3324 max detector TPR @ 1% FPR

Hardest generator config

deepseek-v4.1-flash (Temp: Unknown)

0.4579 max detector TPR @ 1% FPR

336,087

rows

generator configs

prompt subsets

detectors

01Leaderboard 02Model analytics 03Distances 04Appendix

Detector leaderboard

Score-based detectors ranked by overall AUROC. Thresholds are placed exactly on every human score, so TPR is reported at a true 1% and 0.1% false positive rate. Open a row for its threshold sweep, its TPR as AI rows with low cosdist are dropped, and its score distributions.

01Giga EditLens Llama-3.2-3B ScoreN672,174AUROC0.9344TPR @ 1% FPR0.7121TPR @ 0.1% FPR0.5419⌄

1% FPR threshold 0.12600.1% FPR threshold 0.3956

Threshold sweep: Giga EditLens Llama-3.2-3B Score

TPR and FPR at every threshold · both FPR operating points markedOpen full size ↗

Giga EditLens Llama-3.2-3B Score: TPR against minimum cosdist

AI rows with cosdist below the cutoff are dropped · human rows and thresholds are unchangedOpen full size ↗

Score distributions

Giga EditLens Llama-3.2-3B Score: overall human

Overall · human onlyOpen full size ↗

Giga EditLens Llama-3.2-3B Score: overall AI

Overall · AI onlyOpen full size ↗

Giga EditLens Llama-3.2-3B Score: prompt subsets

AI texts · one envelope per prompt subsetOpen full size ↗

Giga EditLens Llama-3.2-3B Score: generator subsets

AI texts · one envelope per generator configOpen full size ↗

02EditLens Llama-3.2-3B ScoreN672,174AUROC0.9189TPR @ 1% FPR0.6763TPR @ 0.1% FPR0.5104⌄

1% FPR threshold 0.28570.1% FPR threshold 0.4744

Threshold sweep: EditLens Llama-3.2-3B Score

TPR and FPR at every threshold · both FPR operating points markedOpen full size ↗

EditLens Llama-3.2-3B Score: TPR against minimum cosdist

AI rows with cosdist below the cutoff are dropped · human rows and thresholds are unchangedOpen full size ↗

Score distributions

EditLens Llama-3.2-3B Score: overall human

Overall · human onlyOpen full size ↗

Overall · AI onlyOpen full size ↗

EditLens Llama-3.2-3B Score: prompt subsets

AI texts · one envelope per prompt subsetOpen full size ↗

EditLens Llama-3.2-3B Score: generator subsets

AI texts · one envelope per generator configOpen full size ↗

03Giga EditLens Roberta ScoreN672,174AUROC0.9113TPR @ 1% FPR0.6156TPR @ 0.1% FPR0.3555⌄

1% FPR threshold 0.38740.1% FPR threshold 0.9789

Threshold sweep: Giga EditLens Roberta Score

TPR and FPR at every threshold · both FPR operating points markedOpen full size ↗

Giga EditLens Roberta Score: TPR against minimum cosdist

AI rows with cosdist below the cutoff are dropped · human rows and thresholds are unchangedOpen full size ↗

Score distributions

Giga EditLens Roberta Score: overall human

Overall · human onlyOpen full size ↗

Overall · AI onlyOpen full size ↗

Giga EditLens Roberta Score: prompt subsets

AI texts · one envelope per prompt subsetOpen full size ↗

Giga EditLens Roberta Score: generator subsets

AI texts · one envelope per generator configOpen full size ↗

04EditLens Roberta-Large ScoreN672,174AUROC0.8904TPR @ 1% FPR0.5051TPR @ 0.1% FPR0.2490⌄

1% FPR threshold 0.44680.1% FPR threshold 0.9694

Threshold sweep: EditLens Roberta-Large Score

TPR and FPR at every threshold · both FPR operating points markedOpen full size ↗

EditLens Roberta-Large Score: TPR against minimum cosdist

AI rows with cosdist below the cutoff are dropped · human rows and thresholds are unchangedOpen full size ↗

Score distributions

EditLens Roberta-Large Score: overall human

Overall · human onlyOpen full size ↗

EditLens Roberta-Large Score: overall AI

Overall · AI onlyOpen full size ↗

EditLens Roberta-Large Score: prompt subsets

AI texts · one envelope per prompt subsetOpen full size ↗

EditLens Roberta-Large Score: generator subsets

AI texts · one envelope per generator configOpen full size ↗

Model-Specific Analytics

Select a model, then a split; the split's charts appear below it. Topic × format compares human scores, AI scores and their unsigned gap; generator × prompt category shows only the AI side, since neither changes the human text; specific prompt ranks every first message by difficulty.

Select a model

01Giga EditLens Llama-3.2-3B ScoreAUROC 0.9344 · TPR@1% 0.7121⌄

Select a split · Giga EditLens Llama-3.2-3B Score

Topic × formathuman · AI · |AI − human| gridsshow ▾

Giga EditLens Llama-3.2-3B Score: Human texts · mean score per topic and format

Human texts · mean score per topic and formatOpen full size ↗

Giga EditLens Llama-3.2-3B Score: AI texts · mean score, same colour scale as the human grid

AI texts · mean score, same colour scale as the human gridOpen full size ↗

Giga EditLens Llama-3.2-3B Score: Mean unsigned |AI − human| score per pair

Mean unsigned |AI − human| score per pairOpen full size ↗

Generator × prompt categoryAI · |AI − human| gridsshow ▾

Giga EditLens Llama-3.2-3B Score: AI texts · mean score per generator config and prompt category

AI texts · mean score per generator config and prompt categoryOpen full size ↗

Mean unsigned |AI − human| score per pairOpen full size ↗

Specific prompt361 prompts · hardest firstshow ▾

One row per distinct first message (a second turn, where there is one, is ignored), hardest first: ranked by TPR at the detector's corpus-wide 1% FPR threshold. Score columns are the detector's mean over the prompt's rows, |AI − human| is the mean per-pair score movement, and distance columns are the mean distance between each source and its rewrite. Only the 12 hardest of 361 are shown; the full ranking ↗ is a CSV beside this card.

Prompt (first message)	Rows	AUROC	TPR @ 1% FPR	TPR @ 0.1% FPR	Human score	AI score	\|AI − human\|	jaccard_1	cosdist
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the topic <<topic>>. rewrite · single-turn	341	0.5714	0.0235	0.0088	0.0106	0.0181	0.0108	0.1284	0.0099
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	323	0.5613	0.0310	0.0000	0.0105	0.0164	0.0118	0.1384	0.0089
Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the topic <<topic>>. rewrite · single-turn	301	0.5890	0.0332	0.0100	0.0087	0.0205	0.0130	0.1763	0.0312
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the format <<format>>. rewrite · single-turn	331	0.5371	0.0332	0.0060	0.0145	0.0186	0.0106	0.1180	0.0069
Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	336	0.5787	0.0476	0.0089	0.0114	0.0219	0.0132	0.1741	0.0295
Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the topic <<topic>>. rewrite · single-turn	378	0.5780	0.0476	0.0185	0.0113	0.0279	0.0187	0.1905	0.0313
Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the format <<format>>. rewrite · single-turn	355	0.6002	0.0535	0.0056	0.0096	0.0246	0.0172	0.1841	0.0282
Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the format <<format>>. rewrite · single-turn	293	0.6280	0.0580	0.0068	0.0120	0.0261	0.0158	0.1740	0.0304
Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	352	0.5857	0.0597	0.0028	0.0111	0.0271	0.0207	0.2002	0.0349
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the format <<format>>. rewrite · single-turn	334	0.6136	0.0599	0.0030	0.0112	0.0269	0.0182	0.1364	0.0096
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the topic <<topic>>. rewrite · single-turn	337	0.6319	0.0653	0.0059	0.0081	0.0300	0.0230	0.1418	0.0080
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	341	0.6459	0.0674	0.0059	0.0112	0.0321	0.0233	0.1432	0.0116

02EditLens Llama-3.2-3B ScoreAUROC 0.9189 · TPR@1% 0.6763⌄

Select a split · EditLens Llama-3.2-3B Score

Topic × formathuman · AI · |AI − human| gridsshow ▾

EditLens Llama-3.2-3B Score: Human texts · mean score per topic and format

Human texts · mean score per topic and formatOpen full size ↗

EditLens Llama-3.2-3B Score: AI texts · mean score, same colour scale as the human grid

AI texts · mean score, same colour scale as the human gridOpen full size ↗

EditLens Llama-3.2-3B Score: Mean unsigned |AI − human| score per pair

Mean unsigned |AI − human| score per pairOpen full size ↗

Generator × prompt categoryAI · |AI − human| gridsshow ▾

EditLens Llama-3.2-3B Score: AI texts · mean score per generator config and prompt category

AI texts · mean score per generator config and prompt categoryOpen full size ↗

Mean unsigned |AI − human| score per pairOpen full size ↗

Specific prompt361 prompts · hardest firstshow ▾

Prompt (first message)	Rows	AUROC	TPR @ 1% FPR	TPR @ 0.1% FPR	Human score	AI score	\|AI − human\|	jaccard_1	cosdist
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the format <<format>>. rewrite · single-turn	331	0.5420	0.0151	0.0000	0.0443	0.0507	0.0211	0.1180	0.0069
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	323	0.5828	0.0155	0.0000	0.0395	0.0562	0.0240	0.1384	0.0089
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the topic <<topic>>. rewrite · single-turn	341	0.5713	0.0235	0.0088	0.0435	0.0545	0.0236	0.1284	0.0099
Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the format <<format>>. rewrite · single-turn	355	0.5774	0.0254	0.0056	0.0418	0.0602	0.0307	0.1841	0.0282
Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	336	0.5748	0.0298	0.0060	0.0410	0.0545	0.0252	0.1741	0.0295
Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the topic <<topic>>. rewrite · single-turn	301	0.5608	0.0299	0.0100	0.0392	0.0509	0.0226	0.1763	0.0312
Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the format <<format>>. rewrite · single-turn	293	0.6101	0.0341	0.0102	0.0414	0.0665	0.0325	0.1740	0.0304
Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	352	0.5763	0.0398	0.0085	0.0427	0.0634	0.0308	0.2002	0.0349
Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the topic <<topic>>. rewrite · single-turn	378	0.5589	0.0423	0.0159	0.0441	0.0609	0.0304	0.1905	0.0313
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the format <<format>>. rewrite · single-turn	334	0.6073	0.0449	0.0030	0.0380	0.0640	0.0388	0.1364	0.0096
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the topic <<topic>>. rewrite · single-turn	337	0.6163	0.0475	0.0089	0.0393	0.0703	0.0434	0.1418	0.0080
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	341	0.6402	0.0499	0.0088	0.0397	0.0806	0.0488	0.1432	0.0116

03Giga EditLens Roberta ScoreAUROC 0.9113 · TPR@1% 0.6156⌄

Select a split · Giga EditLens Roberta Score

Topic × formathuman · AI · |AI − human| gridsshow ▾

Giga EditLens Roberta Score: Human texts · mean score per topic and format

Human texts · mean score per topic and formatOpen full size ↗

Giga EditLens Roberta Score: AI texts · mean score, same colour scale as the human grid

AI texts · mean score, same colour scale as the human gridOpen full size ↗

Giga EditLens Roberta Score: Mean unsigned |AI − human| score per pair

Mean unsigned |AI − human| score per pairOpen full size ↗

Generator × prompt categoryAI · |AI − human| gridsshow ▾

Giga EditLens Roberta Score: AI texts · mean score per generator config and prompt category

AI texts · mean score per generator config and prompt categoryOpen full size ↗

Mean unsigned |AI − human| score per pairOpen full size ↗

Specific prompt361 prompts · hardest firstshow ▾

Prompt (first message)	Rows	AUROC	TPR @ 1% FPR	TPR @ 0.1% FPR	Human score	AI score	\|AI − human\|	jaccard_1	cosdist
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	323	0.5354	0.0124	0.0000	0.0183	0.0302	0.0232	0.1384	0.0089
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the topic <<topic>>. rewrite · single-turn	341	0.5478	0.0176	0.0059	0.0278	0.0362	0.0201	0.1284	0.0099
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the format <<format>>. rewrite · single-turn	334	0.5968	0.0210	0.0000	0.0173	0.0434	0.0347	0.1364	0.0096
Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the topic <<topic>>. rewrite · single-turn	301	0.5749	0.0233	0.0033	0.0163	0.0393	0.0264	0.1763	0.0312
Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	336	0.5721	0.0238	0.0000	0.0177	0.0365	0.0290	0.1741	0.0295
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the format <<format>>. rewrite · single-turn	331	0.5299	0.0242	0.0030	0.0338	0.0376	0.0181	0.1180	0.0069
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the topic <<topic>>. rewrite · single-turn	337	0.6175	0.0267	0.0000	0.0204	0.0475	0.0314	0.1418	0.0080
Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the topic <<topic>>. rewrite · single-turn	378	0.5714	0.0344	0.0000	0.0187	0.0449	0.0323	0.1905	0.0313
Edit this text by combining at least four pairs of short, adjacent sentences into complex sentences using appropriate conjunctions, leaving all other sentences untouched. The final text must have the format <<format>>. rewrite · single-turn	401	0.6348	0.0374	0.0025	0.0205	0.0523	0.0394	0.2006	0.0199
Edit this text by combining at least four pairs of short, adjacent sentences into complex sentences using appropriate conjunctions, leaving all other sentences untouched. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	401	0.6276	0.0374	0.0050	0.0256	0.0577	0.0476	0.2061	0.0242
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	341	0.6195	0.0411	0.0000	0.0228	0.0564	0.0412	0.1432	0.0116
Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	352	0.5597	0.0426	0.0000	0.0307	0.0521	0.0360	0.2002	0.0349

04EditLens Roberta-Large ScoreAUROC 0.8904 · TPR@1% 0.5051⌄

Select a split · EditLens Roberta-Large Score

Topic × formathuman · AI · |AI − human| gridsshow ▾

EditLens Roberta-Large Score: Human texts · mean score per topic and format

Human texts · mean score per topic and formatOpen full size ↗

EditLens Roberta-Large Score: AI texts · mean score, same colour scale as the human grid

AI texts · mean score, same colour scale as the human gridOpen full size ↗

EditLens Roberta-Large Score: Mean unsigned |AI − human| score per pair

Mean unsigned |AI − human| score per pairOpen full size ↗

Generator × prompt categoryAI · |AI − human| gridsshow ▾

EditLens Roberta-Large Score: AI texts · mean score per generator config and prompt category

AI texts · mean score per generator config and prompt categoryOpen full size ↗

Mean unsigned |AI − human| score per pairOpen full size ↗

Specific prompt361 prompts · hardest firstshow ▾

Prompt (first message)	Rows	AUROC	TPR @ 1% FPR	TPR @ 0.1% FPR	Human score	AI score	\|AI − human\|	jaccard_1	cosdist
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the topic <<topic>>. rewrite · single-turn	341	0.5604	0.0088	0.0000	0.0621	0.0637	0.0294	0.1284	0.0099
Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	336	0.5756	0.0149	0.0030	0.0488	0.0663	0.0310	0.1741	0.0295
Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the topic <<topic>>. rewrite · single-turn	301	0.5841	0.0166	0.0033	0.0442	0.0711	0.0313	0.1763	0.0312
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the format <<format>>. rewrite · single-turn	331	0.5507	0.0181	0.0030	0.0744	0.0758	0.0273	0.1180	0.0069
Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	323	0.5803	0.0186	0.0000	0.0576	0.0656	0.0300	0.1384	0.0089
Edit this text by inserting a single, highly engaging introductory sentence at the very beginning, without modifying any of the original text that follows. The final text must have the format <<format>>. rewrite · single-turn	362	0.6395	0.0193	0.0028	0.0603	0.1051	0.0564	0.0858	0.0296
Edit this text by combining at least four pairs of short, adjacent sentences into complex sentences using appropriate conjunctions, leaving all other sentences untouched. The final text must have the format <<format>>. rewrite · single-turn	401	0.6749	0.0200	0.0000	0.0594	0.1112	0.0643	0.2006	0.0199
Edit this text by combining at least four pairs of short, adjacent sentences into complex sentences using appropriate conjunctions, leaving all other sentences untouched. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	401	0.6746	0.0200	0.0100	0.0639	0.1154	0.0716	0.2061	0.0242
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the topic <<topic>>. rewrite · single-turn	337	0.6190	0.0208	0.0000	0.0561	0.0858	0.0419	0.1418	0.0080
Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	352	0.5830	0.0227	0.0000	0.0665	0.0898	0.0423	0.2002	0.0349
Edit this text by replacing, removing or inserting exactly one to three sentences, without changing anything else. The final text must have the format <<format>>. rewrite · single-turn	299	0.6191	0.0234	0.0000	0.0524	0.0942	0.0486	0.1224	0.0241
Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the format <<format>> and topic <<topic>>. rewrite · single-turn	341	0.6245	0.0235	0.0088	0.0526	0.0920	0.0595	0.1432	0.0116

How far do the rewrites move?

Every distance measure between each source and its rewrite: overall, then with every prompt subset and every generator config overlaid on one chart each.

jaccard_1overall · prompt subsets · generator configs⌄

OverallOpen full size ↗

One envelope per prompt subsetOpen full size ↗

One envelope per generator configOpen full size ↗

jaccard_2overall · prompt subsets · generator configs⌄

OverallOpen full size ↗

One envelope per prompt subsetOpen full size ↗

One envelope per generator configOpen full size ↗

levenshteinoverall · prompt subsets · generator configs⌄

OverallOpen full size ↗

One envelope per prompt subsetOpen full size ↗

One envelope per generator configOpen full size ↗

cosdistoverall · prompt subsets · generator configs⌄

OverallOpen full size ↗

One envelope per prompt subsetOpen full size ↗

One envelope per generator configOpen full size ↗

Appendix

Reference material: how the dataset was built and evaluated, its schema and loading code, univariate statistics, and cross-statistic correlations.

How the dataset was built15 generator configs · 4 prompt subsets · evaluation protocol⌄

Every row pairs original (Human) and final_response (AI): a human-written source and the text a generator model produced from it. Inclusion filtering happens while the dataset is built, so this report evaluates the published corpus exactly as it is.

Generator configs

Generator model	Temp	Sampling	Rows	Share
poolside/Laguna-S-2.1-NVFP4	1.0	top_p 1.0 · top_k 20 · enable_thinking False	28,809	8.6%
nvidia/Llama-3.3-70B-Instruct-NVFP4	0.6	top_p 0.9 · enable_thinking False	29,608	8.8%
nvidia/Llama-3.3-70B-Instruct-NVFP4	1.25	top_p 1.0 · top_k -1 · top_a 0.1 · xtc_probability 0.3 · nsigma 1.5 · enable_thinking False	29,397	8.7%
mistralai/Mistral-Small-4-119B-2603-NVFP4	0.7	reasoning_effort none	29,204	8.7%
mistralai/Mistral-Small-4-119B-2603-NVFP4	1.25	top_p 1.0 · reasoning_effort none · top_k -1 · top_a 0.1 · xtc_probability 0.3 · nsigma 1.5	29,619	8.8%
ornith-ai/Ornith-1.5-35B-A3B-NVFP4	0.6	top_p 0.95 · top_k 20 · enable_thinking False	27,801	8.3%
cyankiwi/Qwen3.8-27B-AWQ-INT4	0.7	top_p 0.8 · presence_penalty 1.5 · top_k 20 · enable_thinking False	28,034	8.3%
cyankiwi/Qwen3.8-27B-AWQ-INT4	1.25	top_p 1.0 · top_k -1 · top_a 0.1 · xtc_probability 0.3 · nsigma 1.5 · enable_thinking False	28,923	8.6%
claude-haiku-4-5-20251001	—	type disabled	4,960	1.5%
deepseek/deepseek-v4.1-flash	—	enabled False · max_completion_tokens 16000	4,497	1.3%
cyankiwi/gemma-4-31B-it-AWQ-4bit	1.0	top_p 0.95 · top_k 64 · enable_thinking False	28,750	8.6%
cyankiwi/gemma-4-31B-it-AWQ-4bit	1.25	top_p 1.0 · top_k -1 · top_a 0.1 · xtc_probability 0.3 · nsigma 1.5 · enable_thinking False	28,929	8.6%
gpt-4.1-mini	—	—	4,858	1.4%
gpt-5.6-luna	—	reasoning_effort none	4,831	1.4%
ibm-granite/granite-4.2-30b-nvfp4	1.0	top_p 0.95 · enable_thinking False	27,867	8.3%
Total			336,087	100%

Prompt subsets

Instructions are grouped by their first message. Counts cover all 336,087 rows.

direct_reference85,533 rows

single-turn · 60 instructions

Show the 60 instructions ▾

1,570× Write a different piece of text on the same topic. The final text must have the format <<format>> and topic <<topic>>.
1,567× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should use varied sentence lengths, and remove all common AI buzzwords such as 'tapestry', 'delve', or 'testament', to ensure that it sounds like it was written by a human. The final text must have the topic <<topic>>.
1,539× Write another piece of text that resembles the style and language of this text. The final text must have the topic <<topic>>.
1,515× Write another piece of text that is of a similar theme or background as this text. The final text must have the format <<format>> and topic <<topic>>.
1,513× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should write the new text as if it was a rough draft being typed out on a phone. The final text must have the format <<format>> and topic <<topic>>.
1,505× Write another piece of text that is of a similar theme or background as this text. The final text must have the format <<format>>.
1,500× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should make sure there are slight, subtle grammatical mistakes, to make it sound like it was written by a human. The final text must have the format <<format>> and topic <<topic>>.
1,498× Write a piece of text in the same style and language as this text, but is directly contrasting to the themes, ideas, and/or details in this text. The final text must have the format <<format>> and topic <<topic>>.
1,490× Write a followup to this piece of text, in the same style and language as the text. The final text must have the format <<format>> and topic <<topic>>.
1,489× Write a prequel to this text, describing the events, thoughts, or historical context that immediately led up to this piece being written, making sure to use similar language and wording as the author, as if it was written by the author.
1,486× Write a different piece of text on the same topic.
1,484× Take a subpoint within this text and write a full new text about it, using the language and wording from the text as appropriate. The final text must have the format <<format>> and topic <<topic>>.
1,471× Write a piece of text with a different topic that the author of this text might write about. The final text must have the topic <<topic>>.
1,467× Write a piece of text with a different topic that the author of this text might write about.
1,461× Write a followup to this piece of text, in the same style and language as the text.
1,459× Write a prequel to this text, describing the events, thoughts, or historical context that immediately led up to this piece being written, making sure to use similar language and wording as the author, as if it was written by the author. The final text must have the format <<format>>.
1,458× Write a new piece of text that uses this text as a citation or reference. The final text must have the format <<format>>.
1,457× Write a piece of text in the same style and language as this text, but is directly contrasting to the themes, ideas, and/or details in this text.
1,456× Write another piece of text that resembles the style and language of this text. The final text must have the format <<format>>.
1,451× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should make sure there are slight, subtle grammatical mistakes, to make it sound like it was written by a human. The final text must have the topic <<topic>>.
1,449× Write another piece of text that is of a similar theme or background as this text.
1,448× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should write the new text as if it was a rough draft being typed out on a phone. The final text must have the topic <<topic>>.
1,448× Write a followup to this piece of text, in the same style and language as the text. The final text must have the topic <<topic>>.
1,448× Write a piece of text with a different topic that the author of this text might write about. The final text must have the format <<format>>.
1,448× Write another piece of text that resembles the style and language of this text. The final text must have the format <<format>> and topic <<topic>>.
1,447× Take a subpoint within this text and write a full new text about it, using the language and wording from the text as appropriate. The final text must have the topic <<topic>>.
1,447× Write a detailed rebuttal or critique challenging the core ideas, themes, or facts presented in this text, using the same language and roughly matching the intellectual level of the text. The final text must have the format <<format>> and topic <<topic>>.
1,442× Write a piece of text in the same style and language as this text, but is directly contrasting to the themes, ideas, and/or details in this text. The final text must have the topic <<topic>>.
1,440× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should make sure there are slight, subtle grammatical mistakes, to make it sound like it was written by a human. The final text must have the format <<format>>.
1,434× Write a piece of text with a different topic that the author of this text might write about. The final text must have the format <<format>> and topic <<topic>>.
1,431× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should write the new text as if it was a rough draft being typed out on a phone. The final text must have the format <<format>>.
1,426× Write a new piece of text that uses this text as a citation or reference. The final text must have the format <<format>> and topic <<topic>>.
1,423× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should use varied sentence lengths, and remove all common AI buzzwords such as 'tapestry', 'delve', or 'testament', to ensure that it sounds like it was written by a human.
1,416× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should use varied sentence lengths, and remove all common AI buzzwords such as 'tapestry', 'delve', or 'testament', to ensure that it sounds like it was written by a human. The final text must have the format <<format>> and topic <<topic>>.
1,412× Write a prequel to this text, describing the events, thoughts, or historical context that immediately led up to this piece being written, making sure to use similar language and wording as the author, as if it was written by the author. The final text must have the format <<format>> and topic <<topic>>.
1,412× Write a prequel to this text, describing the events, thoughts, or historical context that immediately led up to this piece being written, making sure to use similar language and wording as the author, as if it was written by the author. The final text must have the topic <<topic>>.
1,410× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should write the new text as if it was a rough draft being typed out on a phone.
1,407× Write a new piece of text that is on a completely different topic, but reuses verbatim as many phrases and sentences from this text as possible.
1,407× Write a piece of text in the same style and language as this text, but is directly contrasting to the themes, ideas, and/or details in this text. The final text must have the format <<format>>.
1,402× Write a different piece of text on the same topic. The final text must have the topic <<topic>>.
1,400× Write a new piece of text that uses this text as a citation or reference. The final text must have the topic <<topic>>.
1,400× Write an alternate version of this text, it should follow the style and format of the original text. The final text must have the format <<format>> and topic <<topic>>.
1,397× Take a subpoint within this text and write a full new text about it, using the language and wording from the text as appropriate. The final text must have the format <<format>>.
1,395× Write a detailed rebuttal or critique challenging the core ideas, themes, or facts presented in this text, using the same language and roughly matching the intellectual level of the text. The final text must have the topic <<topic>>.
1,394× Write a different piece of text on the same topic. The final text must have the format <<format>>.
1,394× Write an alternate version of this text, it should follow the style and format of the original text.
1,383× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should make sure there are slight, subtle grammatical mistakes, to make it sound like it was written by a human.
1,374× Write a detailed rebuttal or critique challenging the core ideas, themes, or facts presented in this text, using the same language and roughly matching the intellectual level of the text. The final text must have the format <<format>>.
1,372× Take a subpoint within this text and write a full new text about it, using the language and wording from the text as appropriate.
1,372× Write a followup to this piece of text, in the same style and language as the text. The final text must have the format <<format>>.
1,366× Take inspiration from the style, language, and content of this text, and write a new piece of text on a topic of your choice. However, you should use varied sentence lengths, and remove all common AI buzzwords such as 'tapestry', 'delve', or 'testament', to ensure that it sounds like it was written by a human. The final text must have the format <<format>>.
1,353× Write an alternate version of this text, it should follow the style and format of the original text. The final text must have the topic <<topic>>.
1,353× Write another piece of text that is of a similar theme or background as this text. The final text must have the topic <<topic>>.
1,346× Write a new piece of text that uses this text as a citation or reference.
1,346× Write another piece of text that resembles the style and language of this text.
1,343× Write a detailed rebuttal or critique challenging the core ideas, themes, or facts presented in this text, using the same language and roughly matching the intellectual level of the text.
1,318× Write an alternate version of this text, it should follow the style and format of the original text. The final text must have the format <<format>>.
1,317× Write a new piece of text that is on a completely different topic, but reuses verbatim as many phrases and sentences from this text as possible. The final text must have the format <<format>> and topic <<topic>>.
1,246× Write a new piece of text that is on a completely different topic, but reuses verbatim as many phrases and sentences from this text as possible. The final text must have the topic <<topic>>.
1,231× Write a new piece of text that is on a completely different topic, but reuses verbatim as many phrases and sentences from this text as possible. The final text must have the format <<format>>.

indirect_reference85,139 rows

two-turn · 25 instructions

Show the 25 instructions ▾

4,391× Translate the given text to German. Do not output anything besides what you were requested to write, and do not output any extra commentary.
4,381× Translate the given text to Spanish. Do not output anything besides what you were requested to write, and do not output any extra commentary.
4,352× Translate the given text to French. Do not output anything besides what you were requested to write, and do not output any extra commentary.
4,319× Translate the given text to Hindi. Do not output anything besides what you were requested to write, and do not output any extra commentary.
4,294× Translate the given text to Chinese. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,643× Convert this text into a series of logical propositions or syllogisms that represent the core argument or narrative progression, stripped of all rhetorical flair. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,621× Take this text, but extract only the most meaningful sentences out of it to create a new text that is a series of disjoint sentences representing the original text. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,609× Create a prompt that might cause an LLM to generate an output resembling this text. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,593× Create a prompt that would allow an LLM to generate this text as accurately/verbatim as possible, without the prompt containing any of the actual text. That is, you must use descriptors to accurately describe the text to generate. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,590× Provide only the first and last word of every sentence in this text, replacing the middle of each sentence with an ellipsis '...'. Maintain the original paragraph structure. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,570× This piece of text was cleverely generated by an LLM with a human supervising it so that the text seems authentic. Reverse engineer the process and come up with a prompt that is most probable to be the one used to generate the text. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,521× Summarize this text, attempting to preserve as much of the original language of the text as possible. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,480× Formulate a set of strict, absolute constraints regarding format, word count, vocabulary restrictions, and syntax rules that, if followed perfectly by an AI, would result in generating a text nearly identical to this one. Note that your response should be formatted as a prompt that would be sent to an AI requesting it to generate something, not just a list of rules. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,418× Write a 'few-shot' prompt using minimal, fragmented excerpts from this text as examples. The prompt should be designed to teach an LLM the pattern needed to generate the rest of the text from scratch. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,335× Take this text and replace one word in every four with a single underscore (use one underscore for each removed word, never several in a row). This way, it will be appropriate for a FIM task. You may decide which words to replace. Reproduce the rest of the document in full, in its original order and paragraph structure. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,309× Replace every verb in this text with the tag [VERB], keeping all nouns, adjectives, articles and punctuation exactly as they appear. Reproduce the document in full, in its original order and paragraph structure. Do not output a legend or a list of the tags you used. Do not output anything besides what you were requested to write, and do not output any extra commentary.
3,245× Create the shortest possible prompt that can be sent to an LLM to reproduce this text. The prompt may include verbatim pieces of the original text, however, it must be short and accurate, working as an effective compression of the original text. Do not output anything besides what you were requested to write, and do not output any extra commentary.
2,714× Write a detailed descriptor for the manner in which the author of this uses language, and things such as their word choice. Do not output anything besides what you were requested to write, and do not output any extra commentary.
2,713× Reformat this text into a sensible, structured JSON object. Do not output anything besides what you were requested to write, and do not output any extra commentary.
2,708× Create a detailed dictionary that fully describes all the meaningful phrases and words this author utilizes to create this text, and how the author utilizes those words and phrases, and why. Do not output anything besides what you were requested to write, and do not output any extra commentary.
2,694× Write a detailed descriptor for how this text is stylistically differentiated from other texts, including grammatical choices and word choices not present in other texts of this kind. Do not output anything besides what you were requested to write, and do not output any extra commentary.
2,675× Write a detailed descriptor for the style of writing of this text, including its qualitative and quantitative properties. Do not output anything besides what you were requested to write, and do not output any extra commentary.
2,666× Envision the scenario in which the author wrote this text. Describe that scenario in exhaustive detail. Do not output anything besides what you were requested to write, and do not output any extra commentary.
2,658× Write a detailed descriptor for the (imaginary) personality of the author who wrote this text, including their livelihood and day to day activities, as well as how that influences their writing. Do not output anything besides what you were requested to write, and do not output any extra commentary.
2,640× Translate the entirety of this text into a sequence of emojis that captures the literal meaning, tone, and narrative progression of the writing. Do not use any letters or words. Do not output anything besides what you were requested to write, and do not output any extra commentary.

revise86,875 rows

43,078 single · 43,797 two-turn · 172 instructions

Show the 172 instructions ▾

1,321× Clarify this text by making the cause-and-effect relationships more obvious.
1,320× Change the tone of this text to be more objective and unbiased.
1,314× Expand this text by adding real-world applications and specific details.
1,309× Clarify this text by paraphrasing it with different vocabulary and varied sentence structures.
1,297× Expand this text by adding backstory and context to enrich understanding.
1,295× Restructure this text to improve the logical flow and argument progression.
1,294× Change the tone of this text to be more academic and scholarly.
1,291× Change the tone of this text to be more direct and bold.
1,291× Restructure this text to ensure a clear introduction, body, and compelling conclusion.
1,288× Change the tone of this text to be more inspiring and motivational.
1,287× Restructure this text to eliminate awkward phrasing and improve the overall rhythm.
1,282× Adapt this text for a lay audience with no technical background.
1,282× Adapt this text for an elderly adult with limited knowledge of modern day affairs.
1,280× Change the tone of this text to be more urgent and compelling.
1,280× Clarify this text by making the argument and main points stronger and easier to understand.
1,279× Change the tone of this text to be more analytical and logical.
1,278× Change the tone of this text to be more conversational and approachable.
1,278× Restructure this text to arrange the points in order of importance.
1,277× Expand this text by adding concrete examples to illustrate the main points.
1,277× Restructure this text to improve the transitions and connections between ideas.
1,276× Adapt this text for social media sharing.
1,274× Clarify this text by defining any terms that might be unclear to the reader.
1,272× Change the tone of this text to be more persuasive and convincing.
1,266× Edit this text to be more concise and powerful.
1,264× Clarify this text by presenting the same information from a fresh angle.
1,263× Clarify this text by eliminating any ambiguous or vague language.
1,257× Change the tone of this text to be more suited for a business audience.
1,257× Restructure this text to read more naturally, as if written by a native English speaker.
1,256× Expand this text by adding vivid imagery and sensory details to bring it to life.
1,253× Change the tone of this text to be more lighthearted and humorous.
1,252× Change the tone of this text to be more polite and diplomatic.
1,251× Change the tone of this text to be more empathetic and understanding.
1,250× Edit this text to trim the length without losing important information or tone.
1,243× Change the tone of this text to be more enthusiastic and energetic.
1,243× Clarify this text by improving the phrasing while preserving its original intent.
1,236× Adapt this text for a 5th-grade reading level.
1,236× Expand this text by adding storytelling elements and dialogue to increase engagement.
1,235× Adapt this text for a teenage audience.
1,234× Change the tone of this text to be more professional and authoritative.
1,231× Restructure this text to group related ideas and include clear section breaks.
1,219× Change the tone of this text to be more relaxed and friendly.
1,214× Edit this text to eliminate redundant and filler words.
1,198× Expand this text by adding supporting information to elaborate on the key points.
307× Change the tone of this text to be more polite and diplomatic. The final text must have the format <<format>>.
301× Expand this text by adding vivid imagery and sensory details to bring it to life. The final text must have the format <<format>>.
294× Clarify this text by improving the phrasing while preserving its original intent. The final text must have the format <<format>>.
294× Expand this text by adding supporting information to elaborate on the key points. The final text must have the format <<format>>.
293× Change the tone of this text to be more relaxed and friendly. The final text must have the format <<format>>.
292× Expand this text by adding supporting information to elaborate on the key points. The final text must have the format <<format>> and topic <<topic>>.
286× Expand this text by adding storytelling elements and dialogue to increase engagement. The final text must have the format <<format>>.
284× Clarify this text by eliminating any ambiguous or vague language. The final text must have the format <<format>> and topic <<topic>>.
284× Edit this text to trim the length without losing important information or tone. The final text must have the topic <<topic>>.
284× Restructure this text to ensure a clear introduction, body, and compelling conclusion. The final text must have the topic <<topic>>.
283× Change the tone of this text to be more suited for a business audience. The final text must have the format <<format>>.
281× Clarify this text by making the argument and main points stronger and easier to understand. The final text must have the format <<format>>.
280× Change the tone of this text to be more enthusiastic and energetic. The final text must have the topic <<topic>>.
278× Change the tone of this text to be more empathetic and understanding. The final text must have the format <<format>> and topic <<topic>>.
278× Restructure this text to read more naturally, as if written by a native English speaker. The final text must have the topic <<topic>>.
277× Adapt this text for a lay audience with no technical background. The final text must have the format <<format>>.
277× Edit this text to be more concise and powerful. The final text must have the format <<format>> and topic <<topic>>.
276× Edit this text to eliminate redundant and filler words. The final text must have the format <<format>> and topic <<topic>>.
273× Change the tone of this text to be more analytical and logical. The final text must have the format <<format>> and topic <<topic>>.
272× Adapt this text for a 5th-grade reading level. The final text must have the format <<format>>.
272× Expand this text by adding concrete examples to illustrate the main points. The final text must have the format <<format>>.
272× Restructure this text to improve the transitions and connections between ideas. The final text must have the format <<format>> and topic <<topic>>.
271× Adapt this text for an elderly adult with limited knowledge of modern day affairs. The final text must have the topic <<topic>>.
270× Change the tone of this text to be more lighthearted and humorous. The final text must have the format <<format>> and topic <<topic>>.
268× Adapt this text for a teenage audience. The final text must have the topic <<topic>>.
267× Change the tone of this text to be more professional and authoritative. The final text must have the topic <<topic>>.
267× Edit this text to trim the length without losing important information or tone. The final text must have the format <<format>> and topic <<topic>>.
266× Change the tone of this text to be more persuasive and convincing. The final text must have the format <<format>> and topic <<topic>>.
266× Edit this text to eliminate redundant and filler words. The final text must have the topic <<topic>>.
264× Adapt this text for a teenage audience. The final text must have the format <<format>> and topic <<topic>>.
264× Adapt this text for social media sharing. The final text must have the format <<format>> and topic <<topic>>.
263× Change the tone of this text to be more direct and bold. The final text must have the format <<format>>.
263× Clarify this text by presenting the same information from a fresh angle. The final text must have the format <<format>> and topic <<topic>>.
261× Adapt this text for a 5th-grade reading level. The final text must have the topic <<topic>>.
261× Change the tone of this text to be more academic and scholarly. The final text must have the format <<format>> and topic <<topic>>.
261× Restructure this text to improve the transitions and connections between ideas. The final text must have the format <<format>>.
260× Adapt this text for a teenage audience. The final text must have the format <<format>>.
260× Clarify this text by improving the phrasing while preserving its original intent. The final text must have the format <<format>> and topic <<topic>>.
259× Change the tone of this text to be more academic and scholarly. The final text must have the topic <<topic>>.
259× Change the tone of this text to be more professional and authoritative. The final text must have the format <<format>> and topic <<topic>>.
259× Change the tone of this text to be more professional and authoritative. The final text must have the format <<format>>.
259× Restructure this text to eliminate awkward phrasing and improve the overall rhythm. The final text must have the format <<format>> and topic <<topic>>.
258× Change the tone of this text to be more relaxed and friendly. The final text must have the format <<format>> and topic <<topic>>.
257× Change the tone of this text to be more lighthearted and humorous. The final text must have the format <<format>>.
257× Change the tone of this text to be more persuasive and convincing. The final text must have the format <<format>>.
257× Restructure this text to ensure a clear introduction, body, and compelling conclusion. The final text must have the format <<format>> and topic <<topic>>.
256× Change the tone of this text to be more enthusiastic and energetic. The final text must have the format <<format>>.
256× Clarify this text by presenting the same information from a fresh angle. The final text must have the format <<format>>.
256× Edit this text to be more concise and powerful. The final text must have the format <<format>>.
256× Expand this text by adding backstory and context to enrich understanding. The final text must have the format <<format>> and topic <<topic>>.
256× Expand this text by adding real-world applications and specific details. The final text must have the topic <<topic>>.
256× Restructure this text to group related ideas and include clear section breaks. The final text must have the format <<format>>.
255× Change the tone of this text to be more objective and unbiased. The final text must have the topic <<topic>>.
255× Change the tone of this text to be more polite and diplomatic. The final text must have the format <<format>> and topic <<topic>>.
254× Change the tone of this text to be more inspiring and motivational. The final text must have the format <<format>>.
254× Clarify this text by improving the phrasing while preserving its original intent. The final text must have the topic <<topic>>.
254× Expand this text by adding concrete examples to illustrate the main points. The final text must have the topic <<topic>>.
253× Clarify this text by making the cause-and-effect relationships more obvious. The final text must have the format <<format>> and topic <<topic>>.
252× Change the tone of this text to be more conversational and approachable. The final text must have the format <<format>>.
252× Expand this text by adding backstory and context to enrich understanding. The final text must have the format <<format>>.
251× Change the tone of this text to be more conversational and approachable. The final text must have the format <<format>> and topic <<topic>>.
251× Change the tone of this text to be more conversational and approachable. The final text must have the topic <<topic>>.
251× Change the tone of this text to be more suited for a business audience. The final text must have the topic <<topic>>.
251× Clarify this text by defining any terms that might be unclear to the reader. The final text must have the topic <<topic>>.
251× Restructure this text to read more naturally, as if written by a native English speaker. The final text must have the format <<format>>.
250× Change the tone of this text to be more relaxed and friendly. The final text must have the topic <<topic>>.
249× Change the tone of this text to be more enthusiastic and energetic. The final text must have the format <<format>> and topic <<topic>>.
249× Clarify this text by paraphrasing it with different vocabulary and varied sentence structures. The final text must have the format <<format>>.
248× Change the tone of this text to be more inspiring and motivational. The final text must have the format <<format>> and topic <<topic>>.
247× Change the tone of this text to be more empathetic and understanding. The final text must have the topic <<topic>>.
247× Restructure this text to eliminate awkward phrasing and improve the overall rhythm. The final text must have the topic <<topic>>.
246× Change the tone of this text to be more lighthearted and humorous. The final text must have the topic <<topic>>.
246× Clarify this text by paraphrasing it with different vocabulary and varied sentence structures. The final text must have the format <<format>> and topic <<topic>>.
246× Restructure this text to improve the logical flow and argument progression. The final text must have the format <<format>> and topic <<topic>>.
245× Adapt this text for an elderly adult with limited knowledge of modern day affairs. The final text must have the format <<format>>.
245× Adapt this text for social media sharing. The final text must have the topic <<topic>>.
245× Change the tone of this text to be more inspiring and motivational. The final text must have the topic <<topic>>.
245× Expand this text by adding supporting information to elaborate on the key points. The final text must have the topic <<topic>>.
244× Change the tone of this text to be more analytical and logical. The final text must have the format <<format>>.
244× Restructure this text to improve the logical flow and argument progression. The final text must have the topic <<topic>>.
243× Expand this text by adding storytelling elements and dialogue to increase engagement. The final text must have the topic <<topic>>.
242× Clarify this text by eliminating any ambiguous or vague language. The final text must have the topic <<topic>>.
242× Edit this text to be more concise and powerful. The final text must have the topic <<topic>>.
242× Expand this text by adding vivid imagery and sensory details to bring it to life. The final text must have the format <<format>> and topic <<topic>>.
241× Adapt this text for a 5th-grade reading level. The final text must have the format <<format>> and topic <<topic>>.
241× Change the tone of this text to be more suited for a business audience. The final text must have the format <<format>> and topic <<topic>>.
241× Change the tone of this text to be more urgent and compelling. The final text must have the topic <<topic>>.
241× Expand this text by adding real-world applications and specific details. The final text must have the format <<format>>.
240× Clarify this text by presenting the same information from a fresh angle. The final text must have the topic <<topic>>.
240× Restructure this text to arrange the points in order of importance. The final text must have the format <<format>> and topic <<topic>>.
240× Restructure this text to group related ideas and include clear section breaks. The final text must have the format <<format>> and topic <<topic>>.
239× Change the tone of this text to be more persuasive and convincing. The final text must have the topic <<topic>>.
239× Expand this text by adding storytelling elements and dialogue to increase engagement. The final text must have the format <<format>> and topic <<topic>>.
239× Restructure this text to improve the logical flow and argument progression. The final text must have the format <<format>>.
239× Restructure this text to read more naturally, as if written by a native English speaker. The final text must have the format <<format>> and topic <<topic>>.
238× Change the tone of this text to be more urgent and compelling. The final text must have the format <<format>> and topic <<topic>>.
238× Clarify this text by making the argument and main points stronger and easier to understand. The final text must have the topic <<topic>>.
238× Restructure this text to group related ideas and include clear section breaks. The final text must have the topic <<topic>>.
237× Adapt this text for an elderly adult with limited knowledge of modern day affairs. The final text must have the format <<format>> and topic <<topic>>.
237× Edit this text to eliminate redundant and filler words. The final text must have the format <<format>>.
236× Clarify this text by making the cause-and-effect relationships more obvious. The final text must have the format <<format>>.
234× Expand this text by adding backstory and context to enrich understanding. The final text must have the topic <<topic>>.
233× Change the tone of this text to be more urgent and compelling. The final text must have the format <<format>>.
233× Clarify this text by defining any terms that might be unclear to the reader. The final text must have the format <<format>>.
233× Expand this text by adding concrete examples to illustrate the main points. The final text must have the format <<format>> and topic <<topic>>.
233× Expand this text by adding real-world applications and specific details. The final text must have the format <<format>> and topic <<topic>>.
233× Restructure this text to eliminate awkward phrasing and improve the overall rhythm. The final text must have the format <<format>>.
232× Clarify this text by making the argument and main points stronger and easier to understand. The final text must have the format <<format>> and topic <<topic>>.
231× Change the tone of this text to be more direct and bold. The final text must have the topic <<topic>>.
230× Change the tone of this text to be more direct and bold. The final text must have the format <<format>> and topic <<topic>>.
229× Change the tone of this text to be more empathetic and understanding. The final text must have the format <<format>>.
228× Adapt this text for a lay audience with no technical background. The final text must have the topic <<topic>>.
227× Change the tone of this text to be more objective and unbiased. The final text must have the format <<format>> and topic <<topic>>.
227× Change the tone of this text to be more objective and unbiased. The final text must have the format <<format>>.
227× Clarify this text by defining any terms that might be unclear to the reader. The final text must have the format <<format>> and topic <<topic>>.
226× Clarify this text by making the cause-and-effect relationships more obvious. The final text must have the topic <<topic>>.
225× Change the tone of this text to be more analytical and logical. The final text must have the topic <<topic>>.
225× Clarify this text by paraphrasing it with different vocabulary and varied sentence structures. The final text must have the topic <<topic>>.
225× Restructure this text to ensure a clear introduction, body, and compelling conclusion. The final text must have the format <<format>>.
224× Change the tone of this text to be more polite and diplomatic. The final text must have the topic <<topic>>.
223× Adapt this text for a lay audience with no technical background. The final text must have the format <<format>> and topic <<topic>>.
222× Restructure this text to arrange the points in order of importance. The final text must have the format <<format>>.
221× Edit this text to trim the length without losing important information or tone. The final text must have the format <<format>>.
220× Restructure this text to improve the transitions and connections between ideas. The final text must have the topic <<topic>>.
218× Clarify this text by eliminating any ambiguous or vague language. The final text must have the format <<format>>.
216× Adapt this text for social media sharing. The final text must have the format <<format>>.
210× Change the tone of this text to be more academic and scholarly. The final text must have the format <<format>>.
206× Expand this text by adding vivid imagery and sensory details to bring it to life. The final text must have the topic <<topic>>.
200× Restructure this text to arrange the points in order of importance. The final text must have the topic <<topic>>.

rewrite78,540 rows

36,574 single · 41,966 two-turn · 104 instructions

Show the 104 instructions ▾

2,118× Rewrite only the opening hook and the final concluding sentence of this text, leaving the entire middle portion exactly as it was originally written.
2,084× Edit this text by replacing the final sentence with a more open-ended, thought-provoking statement, without altering any preceding text.
2,073× Modify this text to be as long as possible. At least half of the original text must be repeated verbatim, and at most three quarters of it; the remainder must be new prose.
2,059× Edit this text by completely rewriting the topic sentence of every paragraph, while preserving the supporting sentences in those paragraphs verbatim.
2,058× Edit this text by inserting a singular new section (in an appropriate locations), you may only make minimal modifications to transitions in adjacent sections for better flow.
2,048× Edit this text by upgrading the verbs to be more dynamic and action-oriented, leaving the subjects, objects, and overall length completely unchanged.
2,043× Improve this text by rewriting it. At least half of the original text must be repeated verbatim, and at most three quarters of it; the remainder must be new prose.
2,043× Modify this text so that it is as short as possible. At least a quarter of the original text must be repeated verbatim, and at most half of it; the remainder must be new prose.
2,042× Rewrite exactly one section of this text so that its wording is entirely new, and leave every other section untouched. The rewritten section must share no sentence with the original.
2,027× Rewrite the weakest section in this text, but do not make any modifications to any other section.
2,003× Edit this text by combining at least four pairs of short, adjacent sentences into complex sentences using appropriate conjunctions, leaving all other sentences untouched.
1,966× Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow.
1,961× Edit this text by replacing all instances of passive voice with active voice, without altering any sentences that are already in the active voice.
1,924× Edit this text by replacing five to ten percent of words in it, without changing anything else.
1,918× Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound.
1,911× Edit this text by inserting a single, highly engaging introductory sentence at the very beginning, without modifying any of the original text that follows.
1,908× Edit this text by inserting exactly one to three sentences (in appropriate locations), you may only make minimal modifications to transitions in adjacent sentences for better flow.
1,899× Edit this text by replacing all existing transition words and phrases (e.g., 'however', 'therefore', 'in addition') with different, synonymous transitions, making no other changes.
1,888× Edit this text by replacing generic adjectives with highly specific, evocative ones, modifying no more than ten percent of the total words.
1,862× Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow.
1,862× Edit this text by removing five to ten percent of the words from it (the least necessary), without changing anything else.
1,848× Edit this text by inserting twenty to thirty words (in appropriate locations), without changing anything else.
1,846× Edit this text by replacing, removing or inserting twenty to thirty words, without changing anything else.
1,818× Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text.
1,807× Edit this text by replacing, removing or inserting exactly one to three sentences, without changing anything else.
1,782× Edit this text by replacing exactly one to three sentences, without changing anything else.
491× Improve this text by rewriting it. At least half of the original text must be repeated verbatim, and at most three quarters of it; the remainder must be new prose. The final text must have the topic <<topic>>.
453× Modify this text to be as long as possible. At least half of the original text must be repeated verbatim, and at most three quarters of it; the remainder must be new prose. The final text must have the format <<format>> and topic <<topic>>.
435× Rewrite exactly one section of this text so that its wording is entirely new, and leave every other section untouched. The rewritten section must share no sentence with the original. The final text must have the topic <<topic>>.
429× Modify this text so that it is as short as possible. At least a quarter of the original text must be repeated verbatim, and at most half of it; the remainder must be new prose. The final text must have the topic <<topic>>.
429× Modify this text to be as long as possible. At least half of the original text must be repeated verbatim, and at most three quarters of it; the remainder must be new prose. The final text must have the format <<format>>.
427× Modify this text so that it is as short as possible. At least a quarter of the original text must be repeated verbatim, and at most half of it; the remainder must be new prose. The final text must have the format <<format>>.
425× Edit this text by completely rewriting the topic sentence of every paragraph, while preserving the supporting sentences in those paragraphs verbatim. The final text must have the topic <<topic>>.
425× Edit this text by inserting a singular new section (in an appropriate locations), you may only make minimal modifications to transitions in adjacent sections for better flow. The final text must have the format <<format>> and topic <<topic>>.
424× Edit this text by removing five to ten percent of the words from it (the least necessary), without changing anything else. The final text must have the format <<format>> and topic <<topic>>.
416× Modify this text so that it is as short as possible. At least a quarter of the original text must be repeated verbatim, and at most half of it; the remainder must be new prose. The final text must have the format <<format>> and topic <<topic>>.
411× Edit this text by completely rewriting the topic sentence of every paragraph, while preserving the supporting sentences in those paragraphs verbatim. The final text must have the format <<format>> and topic <<topic>>.
404× Rewrite only the opening hook and the final concluding sentence of this text, leaving the entire middle portion exactly as it was originally written. The final text must have the topic <<topic>>.
401× Edit this text by combining at least four pairs of short, adjacent sentences into complex sentences using appropriate conjunctions, leaving all other sentences untouched. The final text must have the format <<format>> and topic <<topic>>.
401× Edit this text by combining at least four pairs of short, adjacent sentences into complex sentences using appropriate conjunctions, leaving all other sentences untouched. The final text must have the format <<format>>.
394× Edit this text by replacing all instances of passive voice with active voice, without altering any sentences that are already in the active voice. The final text must have the topic <<topic>>.
393× Edit this text by completely rewriting the topic sentence of every paragraph, while preserving the supporting sentences in those paragraphs verbatim. The final text must have the format <<format>>.
393× Edit this text by inserting a singular new section (in an appropriate locations), you may only make minimal modifications to transitions in adjacent sections for better flow. The final text must have the format <<format>>.
393× Rewrite only the opening hook and the final concluding sentence of this text, leaving the entire middle portion exactly as it was originally written. The final text must have the format <<format>> and topic <<topic>>.
391× Rewrite exactly one section of this text so that its wording is entirely new, and leave every other section untouched. The rewritten section must share no sentence with the original. The final text must have the format <<format>> and topic <<topic>>.
388× Improve this text by rewriting it. At least half of the original text must be repeated verbatim, and at most three quarters of it; the remainder must be new prose. The final text must have the format <<format>> and topic <<topic>>.
387× Modify this text to be as long as possible. At least half of the original text must be repeated verbatim, and at most three quarters of it; the remainder must be new prose. The final text must have the topic <<topic>>.
386× Rewrite exactly one section of this text so that its wording is entirely new, and leave every other section untouched. The rewritten section must share no sentence with the original. The final text must have the format <<format>>.
385× Edit this text by replacing the final sentence with a more open-ended, thought-provoking statement, without altering any preceding text. The final text must have the topic <<topic>>.
383× Edit this text by combining at least four pairs of short, adjacent sentences into complex sentences using appropriate conjunctions, leaving all other sentences untouched. The final text must have the topic <<topic>>.
382× Edit this text by upgrading the verbs to be more dynamic and action-oriented, leaving the subjects, objects, and overall length completely unchanged. The final text must have the format <<format>>.
380× Edit this text by inserting a singular new section (in an appropriate locations), you may only make minimal modifications to transitions in adjacent sections for better flow. The final text must have the topic <<topic>>.
380× Edit this text by replacing all instances of passive voice with active voice, without altering any sentences that are already in the active voice. The final text must have the format <<format>>.
379× Edit this text by removing five to ten percent of the words from it (the least necessary), without changing anything else. The final text must have the format <<format>>.
378× Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the topic <<topic>>.
373× Edit this text by inserting exactly one to three sentences (in appropriate locations), you may only make minimal modifications to transitions in adjacent sentences for better flow. The final text must have the topic <<topic>>.
373× Rewrite the weakest section in this text, but do not make any modifications to any other section. The final text must have the format <<format>> and topic <<topic>>.
369× Edit this text by replacing the final sentence with a more open-ended, thought-provoking statement, without altering any preceding text. The final text must have the format <<format>> and topic <<topic>>.
369× Rewrite only the opening hook and the final concluding sentence of this text, leaving the entire middle portion exactly as it was originally written. The final text must have the format <<format>>.
368× Edit this text by inserting exactly one to three sentences (in appropriate locations), you may only make minimal modifications to transitions in adjacent sentences for better flow. The final text must have the format <<format>> and topic <<topic>>.
367× Edit this text by replacing five to ten percent of words in it, without changing anything else. The final text must have the format <<format>> and topic <<topic>>.
364× Edit this text by inserting a single, highly engaging introductory sentence at the very beginning, without modifying any of the original text that follows. The final text must have the topic <<topic>>.
362× Edit this text by inserting a single, highly engaging introductory sentence at the very beginning, without modifying any of the original text that follows. The final text must have the format <<format>>.
361× Edit this text by removing five to ten percent of the words from it (the least necessary), without changing anything else. The final text must have the topic <<topic>>.
359× Edit this text by upgrading the verbs to be more dynamic and action-oriented, leaving the subjects, objects, and overall length completely unchanged. The final text must have the format <<format>> and topic <<topic>>.
359× Rewrite the weakest section in this text, but do not make any modifications to any other section. The final text must have the format <<format>>.
355× Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the format <<format>>.
354× Edit this text by replacing the final sentence with a more open-ended, thought-provoking statement, without altering any preceding text. The final text must have the format <<format>>.
354× Rewrite the weakest section in this text, but do not make any modifications to any other section. The final text must have the topic <<topic>>.
352× Edit this text by removing one section from it (the least necessary), you may only make minimal modifications to adjacent sections for better flow. The final text must have the format <<format>> and topic <<topic>>.
348× Edit this text by upgrading the verbs to be more dynamic and action-oriented, leaving the subjects, objects, and overall length completely unchanged. The final text must have the topic <<topic>>.
347× Improve this text by rewriting it. At least half of the original text must be repeated verbatim, and at most three quarters of it; the remainder must be new prose. The final text must have the format <<format>>.
341× Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the format <<format>> and topic <<topic>>.
341× Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the topic <<topic>>.
337× Edit this text by inserting a single, highly engaging introductory sentence at the very beginning, without modifying any of the original text that follows. The final text must have the format <<format>> and topic <<topic>>.
337× Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the topic <<topic>>.
336× Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the format <<format>> and topic <<topic>>.
334× Edit this text by removing all adverbs, making only minimal adjustments to the surrounding grammar to ensure the sentences remain structurally sound. The final text must have the format <<format>>.
331× Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the format <<format>>.
330× Edit this text by replacing all existing transition words and phrases (e.g., 'however', 'therefore', 'in addition') with different, synonymous transitions, making no other changes. The final text must have the format <<format>> and topic <<topic>>.
328× Edit this text by replacing exactly one to three sentences, without changing anything else. The final text must have the topic <<topic>>.
323× Edit this text by inserting exactly one to three sentences (in appropriate locations), you may only make minimal modifications to transitions in adjacent sentences for better flow. The final text must have the format <<format>>.
323× Edit this text by splitting the three longest sentences into shorter, distinct sentences, without changing the vocabulary or structure of the rest of the text. The final text must have the format <<format>> and topic <<topic>>.
318× Edit this text by inserting twenty to thirty words (in appropriate locations), without changing anything else. The final text must have the format <<format>> and topic <<topic>>.
317× Edit this text by replacing exactly one to three sentences, without changing anything else. The final text must have the format <<format>> and topic <<topic>>.
313× Edit this text by replacing generic adjectives with highly specific, evocative ones, modifying no more than ten percent of the total words. The final text must have the format <<format>> and topic <<topic>>.
313× Edit this text by replacing generic adjectives with highly specific, evocative ones, modifying no more than ten percent of the total words. The final text must have the format <<format>>.
310× Edit this text by replacing five to ten percent of words in it, without changing anything else. The final text must have the format <<format>>.
306× Edit this text by inserting twenty to thirty words (in appropriate locations), without changing anything else. The final text must have the format <<format>>.
303× Edit this text by replacing all instances of passive voice with active voice, without altering any sentences that are already in the active voice. The final text must have the format <<format>> and topic <<topic>>.
303× Edit this text by replacing generic adjectives with highly specific, evocative ones, modifying no more than ten percent of the total words. The final text must have the topic <<topic>>.
301× Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the topic <<topic>>.
299× Edit this text by replacing, removing or inserting exactly one to three sentences, without changing anything else. The final text must have the format <<format>>.
298× Edit this text by replacing, removing or inserting exactly one to three sentences, without changing anything else. The final text must have the topic <<topic>>.
295× Edit this text by replacing exactly one to three sentences, without changing anything else. The final text must have the format <<format>>.
293× Edit this text by inserting twenty to thirty words (in appropriate locations), without changing anything else. The final text must have the topic <<topic>>.
293× Edit this text by removing exactly one to three sentences from it (the least necessary), you may only make minimal modifications to adjacent sentences for better flow. The final text must have the format <<format>>.
291× Edit this text by replacing all existing transition words and phrases (e.g., 'however', 'therefore', 'in addition') with different, synonymous transitions, making no other changes. The final text must have the format <<format>>.
289× Edit this text by replacing, removing or inserting twenty to thirty words, without changing anything else. The final text must have the format <<format>>.
288× Edit this text by replacing all existing transition words and phrases (e.g., 'however', 'therefore', 'in addition') with different, synonymous transitions, making no other changes. The final text must have the topic <<topic>>.
288× Edit this text by replacing five to ten percent of words in it, without changing anything else. The final text must have the topic <<topic>>.
270× Edit this text by replacing, removing or inserting twenty to thirty words, without changing anything else. The final text must have the topic <<topic>>.
251× Edit this text by replacing, removing or inserting twenty to thirty words, without changing anything else. The final text must have the format <<format>> and topic <<topic>>.
245× Edit this text by replacing, removing or inserting exactly one to three sentences, without changing anything else. The final text must have the format <<format>> and topic <<topic>>.

Topics

Topic	Rows	Share
Finance & Business	20,069	6.0%
Health	19,701	5.9%
Food & Dining	17,732	5.3%
Sports & Fitness	17,297	5.1%
Home & Hobbies	17,081	5.1%
Crime & Law	16,662	5.0%
Transportation	16,448	4.9%
Religion	16,331	4.9%
Politics	16,074	4.8%
Education & Jobs	15,886	4.7%
Entertainment	15,813	4.7%
Science & Tech.	15,493	4.6%
Art & Design	14,917	4.4%
Literature	14,687	4.4%
Social Life	14,072	4.2%
Games	14,063	4.2%
Software	12,900	3.8%
Travel	12,680	3.8%
Fashion & Beauty	11,641	3.5%
History	10,141	3.0%
Industrial	9,801	2.9%
Hardware	9,269	2.8%
Software Dev.	7,329	2.2%

Formats

Format	Rows	Share
Personal Blog	40,242	12.0%
News Article	34,868	10.4%
Tutorial	33,432	9.9%
Nonfiction Writing	32,896	9.8%
Product Page	30,930	9.2%
Knowledge Article	29,101	8.7%
Listicle	22,728	6.8%
Comment Section	19,246	5.7%
News (Org.)	17,723	5.3%
About (Org.)	13,097	3.9%
User Review	11,128	3.3%
Audio Transcript	10,272	3.1%
Creative Writing	8,904	2.6%
About (Pers.)	8,392	2.5%
Customer Support	6,592	2.0%
Academic Writing	5,911	1.8%
Q&A Forum	4,582	1.4%
FAQ	3,007	0.9%
Legal Notices	2,180	0.6%
Documentation	789	0.2%
Spam / Ads	67	0.0%

Evaluation protocol

Classes

original is Human, final_response is AI. Each detector scores both texts of every row.

Thresholds

All 336,087 rows are used; there is no sweep/evaluation split. Each detector's thresholds are placed exactly on its human scores so that 1% and 0.1% of human texts are called AI.

Metrics

N (texts scored), AUROC, and TPR at the 1% and 0.1% FPR thresholds. Subsets are scored at the corpus-wide thresholds rather than refitted.

Filters

None. No cosine, soft-ngram, Jaccard, or other analysis-time filtering is applied.

Skipped

Perplexity (Llama-3.2-3B-Instruct) (original_perplexity_llama_instruct; final_response_perplexity_llama_instruct), Perplexity (Llama-3.2-3B) (original_perplexity_llama_base; final_response_perplexity_llama_base), Entropy (Llama-3.2-3B-Instruct) (original_entropy_llama_instruct; final_response_entropy_llama_instruct), Entropy (Llama-3.2-3B) (original_entropy_llama_base; final_response_entropy_llama_base), Top-p Outliers (Llama-3.2-3B-Instruct) (original_topp_outlier_llama_instruct; final_response_topp_outlier_llama_instruct), Top-p Outliers (Llama-3.2-3B) (original_topp_outlier_llama_base; final_response_topp_outlier_llama_base), Top-k Outliers (Llama-3.2-3B-Instruct) (original_topk_outlier_llama_instruct; final_response_topk_outlier_llama_instruct), Top-k Outliers (Llama-3.2-3B) (original_topk_outlier_llama_base; final_response_topk_outlier_llama_base), FastDetectGPT (Llama-3.2-3B-Instruct) (original_fastdetectgpt_llama_instruct; final_response_fastdetectgpt_llama_instruct), FastDetectGPT (Llama-3.2-3B) (original_fastdetectgpt_llama_base; final_response_fastdetectgpt_llama_base), Binoculars (original_binoculars; final_response_binoculars)

Config

config/analysis.toml

Schema & usage29 columns · loading snippet⌄

29 columns per row. Detector columns come in pairs, one per text of the row.

Text

original

Human-written source (class: Human)

final_response

Generated text (class: AI)

response_0

Generator output for the first turn

response_1

Generator output for the second turn (empty for single-turn prompts)

Generation

prompt

struct: chat_turns (list of strings), use_multiturn, examples, metadata.PROMPT_TYPE

generator_model

Hugging Face id of the generator model

generation_params

JSON string of sampling parameters

Metadata

topic

Topic of the human source

format

Format of the human source

Distances

jaccard_1, jaccard_2, levenshtein, cosdist

4 float columns measuring how far the rewrite moved from the original

Other columns

original_editlens_bucket_roberta_large, final_response_editlens_bucket_roberta_large, original_editlens_bucket_llama_3_2_3b, final_response_editlens_bucket_llama_3_2_3b, original_editlens_bucket_giga_roberta, final_response_editlens_bucket_giga_roberta, original_editlens_bucket_giga_llama_3_2_3b, final_response_editlens_bucket_giga_llama_3_2_3b

Not used by this report

Detector scores

One float column per text: {original|final_response}<suffix>

_editlens_score_roberta_large

EditLens Roberta-Large Score

_editlens_score_llama_3_2_3b

EditLens Llama-3.2-3B Score

_editlens_score_giga_roberta

Giga EditLens Roberta Score

_editlens_score_giga_llama_3_2_3b

Giga EditLens Llama-3.2-3B Score

Load it

from datasets import concatenate_datasets, get_dataset_config_names, load_dataset

repo = "G-reen/fastdetector-train-stat-train"
# every generator config is its own dataset config
ds = concatenate_datasets([
    load_dataset(repo, config, split="train")
    for config in get_dataset_config_names(repo)
])

Univariate statistics12 statistics · all 336,087 rows valid⌄

Every statistic the report does arithmetic on, over the complete corpus. Missing or non-finite values are counted as invalid.

Statistic	Mean	Median	Std	Min	Max
`jaccard_1`	0.6287	0.7452	0.2735	0.0000	1.0000
`jaccard_2`	0.7451	0.8997	0.2951	0.0164	1.0000
`levenshtein`	2192.5230	1411.0000	3257.0562	1.0000	175935.0000
`cosdist`	0.1935	0.1307	0.1929	-0.0070	1.0490
`EditLens Roberta-Large Score (Human)`	0.0585	0.0208	0.0972	0.0062	0.9996
`EditLens Roberta-Large Score (AI)`	0.5139	0.4533	0.3730	0.0063	0.9996
`EditLens Llama-3.2-3B Score (Human)`	0.0431	0.0216	0.0585	0.0001	0.9230
`EditLens Llama-3.2-3B Score (AI)`	0.5144	0.4868	0.3520	0.0003	1.0000
`Giga EditLens Roberta Score (Human)`	0.0200	0.0046	0.0762	0.0013	1.0000
`Giga EditLens Roberta Score (AI)`	0.5826	0.7049	0.4145	0.0013	1.0000
`Giga EditLens Llama-3.2-3B Score (Human)`	0.0096	0.0034	0.0295	0.0003	0.9971
`Giga EditLens Llama-3.2-3B Score (AI)`	0.5050	0.4670	0.3941	0.0003	1.0000

Correlation heatmappairwise Pearson r⌄

Each pair is correlated over the rows where both are finite.Open full size ↗

Generated by the FastDetector analysis pipeline from the complete published corpus.config/analysis.toml

Downloads last month: 1,288