koyena/DeepSeek-R1-Distill-Llama-8B-max-activation-SAE-cache-L23 Viewer • Updated May 8, 2025 • 32.8k • 30
koyena/DeepSeek-R1-Distill-Llama-8B-max-activation-SAE-cache-L7 Viewer • Updated May 8, 2025 • 32.8k • 35 • 1
koyena/DeepSeek-R1-Distill-Llama-8B-max-activation-SAE-cache-L7 Viewer • Updated May 8, 2025 • 32.8k • 35 • 1
koyena/DeepSeek-R1-Distill-Llama-8B-max-activation-SAE-cache-L23 Viewer • Updated May 8, 2025 • 32.8k • 30
koyena/Magpie-Reasoning-V2-250K-CoT-Deepseek-R1-Llama-70B-formatted Viewer • Updated Apr 23, 2025 • 250k • 21
koyena/Magpie-Reasoning-V2-250K-CoT-Deepseek-R1-Llama-70B-formatted Viewer • Updated Apr 23, 2025 • 250k • 21
SliderSpace: Decomposing the Visual Capabilities of Diffusion Models Paper • 2502.01639 • Published Feb 3, 2025 • 26
LLMs Know More Than They Show: On the Intrinsic Representation of LLM Hallucinations Paper • 2410.02707 • Published Oct 3, 2024 • 47
NNsight and NDIF: Democratizing Access to Foundation Model Internals Paper • 2407.14561 • Published Jul 18, 2024 • 35
NNsight and NDIF: Democratizing Access to Foundation Model Internals Paper • 2407.14561 • Published Jul 18, 2024 • 35