AI RESEARCH
Select to Think: Unlocking SLM Potential with Local Sufficiency
arXiv CS.CL
•
ArXi:2604.26940v1 Announce Type: new Small language models (SLMs) offer computational efficiency for scalable deployment, yet they often fall short of the reasoning power exhibited by their larger counterparts (LLMs). To mitigate this gap, current approaches invoke an LLM to generate tokens at points of reasoning divergence, but these external calls