AI RESEARCH
DenseSteer: Steering Small Language Models towards Dense Math Reasoning
arXiv CS.AI
•
ArXi:2605.29247v1 Announce Type: new Large language models (LLMs) nstrate strong chain-of-thought (CoT) reasoning abilities, while smaller models (<= 3B parameters) significantly underperform on multi-step reasoning tasks. Based on empirical analyses of the Qwen-2.5 model family on math reasoning benchmarks, we find that proficient reasoning is associated with fewer reasoning steps but higher information density per step, a property we term Dense Reasoning. Motivated by this observation, we propose DenseSteer, a.