3 papers
cs.LG2026
DRIFT: Difficulty Routing Self-DIstillation with Rhythm-Gated Exploration and Success BuFfer Training
Haisen Luo, Yiwei Liu, Haoning Wang +13
Enabling large language models to achieve stable self-improvement without external expert supervision remains a central challenge in complex reasoning tasks. Existing self-distilla…
cs.CL2026
FormulaReasoning: A Dataset for Formula-Based Numerical Reasoning
Xiao Li, Bolin Zhu, Kaiwen Shi +4
The application of physics formulas is a fundamental human capability in numerical reasoning. While existing datasets often rely on implicit mathematical knowledge, they rarely exp…
cs.CL2025
LogiNumSynth: Synthesizing Joint Logical-Numerical Reasoning Problems for Language Models
Yiwei Liu, Yucheng Li, Xiao Li +1
Joint logical-numerical reasoning remains a major challenge for language models, yet existing datasets rely on fixed rule sets and offer limited control over task complexity, const…