Thinking by Subtraction: Confidence-Driven Contrastive Decoding for LLM Reasoning
arXiv:2602.18232v1 Announce Type: new Abstract: Recent work on test-time scaling for large language model (LLM) reasoning typically assumes that allocating more inference-time computation uniformly improves …
Lexiang Tang, Weihao Gao, Bingchen Zhao, Lu Ma, Qiao jin, Bang Yang, Yuexian Zou
3 views
