Garbage Attention in Large Language Models: BOS Sink Heads and Sink-aware Pruning
Paper • 2601.06787 • Published • 1
Sok Jaewon
Jaewon4
AI & ML interests
None yet
Recent Activity
authored a paper 1 day ago
EpiCaR: Knowing What You Don't Know Matters for Better Reasoning in LLMs authored a paper 1 day ago
From Noise to Diversity: Random Embedding Injection in LLM ReasoningOrganizations
None yet
upvoted a paper 1 day ago