GHOST 2.0: generative high-fidelity one shot transfer of heads Paper • 2502.18417 • Published 16 days ago • 63
LLM-Microscope: Uncovering the Hidden Role of Punctuation in Context Memory of Transformers Paper • 2502.15007 • Published 20 days ago • 162
SynthDetoxM: Modern LLMs are Few-Shot Parallel Detoxification Data Annotators Paper • 2502.06394 • Published about 1 month ago • 86
The Differences Between Direct Alignment Algorithms are a Blur Paper • 2502.01237 • Published Feb 3 • 112