Your Transformer Can Hold Two Thoughts at Once: Evidence of Linear Superposition in LLMs
Paper • 2609.29845 • Published • 91
Anton Razzhigaev
razzant
AI & ML interests
Language models, multimodal models, knowledge graphs, chatbots
Recent Activity
authored a paper 4 days ago
Your Transformer Can Hold Two Thoughts at Once: Evidence of Linear Superposition in LLMs submitted a paper 4 days ago
Your Transformer Can Hold Two Thoughts at Once: Evidence of Linear Superposition in LLMsOrganizations
upvoted a paper 4 days ago
upvoted a collection about 1 month ago
upvoted a paper about 2 months ago
upvoted a collection 10 months ago
upvoted 3
papers over 1 year ago
Image Reconstruction as a Tool for Feature Analysis
Paper • 2506.07803 • Published • 29
I Have Covered All the Bases Here: Interpreting Reasoning Features in Large Language Models via Sparse Autoencoders
Paper • 2503.18878 • Published • 121
Entropy-Guided Attention for Private LLMs
Paper • 2501.03489 • Published • 14
upvoted a paper about 2 years ago
upvoted 2
papers over 2 years ago