12 papers
EasyOPD: An Easy-to-use On-Policy Distillation Framework for Large Language Models
Jie Sun, Mao Zheng, Mingyang Song +7
The paper introduces EasyOPD, a modular framework that simplifies on-policy distillation for large language models by separating configuration, supervision logic, and distributed e…
PhysGuard: Fisher-Guided Gradient Projection for Sim-to-Real Neural PDE Surrogates
Changjian Zhou, Junfeng Fang, Negin Yousefpour +3
Neural operator models trained on simulation data often lose accuracy when applied to experimental measurements due to the sim-to-real gap. Standard fine-tuning with limited real d…
On-Policy Distillation with Curriculum Turn-level Guidance for Multi-turn Agents
Gengsheng Li, Mao Zheng, Mingyang Song +8
Multi-turn agents that plan, invoke tools, and interact with environments offer a promising paradigm for solving complex tasks, yet their capabilities typically rely on very large…
Cosine Misleads: Auxiliary Losses Reshape Vision Language Models, Not Their Latents
XiuYu Zhang, Junfeng Fang, Zhenkai Liang
Latent visual reasoning (LVR) inserts supervised latent tokens between perception and answer generation in vision-language models (VLMs). The field uses alignment between these lat…
Self-Evaluation Is Already There: Eliciting Latent Judge Calibration in Base LLMs with Minimal Data
XiuYu Zhang, Yi Shan, Junfeng Fang +1
Large language models are increasingly evaluated by other models, raising a natural question: can a model predict how a judge will score its own output? We find that the ability is…
Learning to Foresee: Unveiling the Unlocking Efficiency of On-Policy Distillation
Yuchen Cai, Ding Cao, Liang Lin +9
On-policy distillation (OPD) has emerged as an efficient post-training paradigm for large language models. However, existing studies largely attribute this advantage to denser and…