Learning to Reason Efficiently with A* Post-Training 文章

ArXiv CS.CL2026-05-26NEWSen作者: Andreas Opedal, Francesco Ignazio Re, Abulhair Saparov, Mrinmaya Sachan, Bernhard Sch\"olkopf, Ryan Cotterell
大语言模型
查看原文 →
关系图谱

Learning to Reason Efficiently with A* Post-Training · 相关技术

相关技术

ORMODELLMlanguage model远程代码执行（RCE）后训练reinforcement learningexecUCTUARTEEStraight-Through EstimatorStanSupervised Fine-TuningReferring expression comprehension (REC)Parts-of-Speech (POS) tagsNatural Language InferenceNATLMMlarge language modelsForFFIEFIARGANNLLaMA