paper-with-me

Papers

On-Policy Delta Distillation for Multilingual Math Reasoning

2026-08-06 · Byeongho Heo, Jaehui Hwang, Sangdoo Yun, Dongyoon Han hf

On-Policy Distillation (OPD) is emerging as a promising alternative to reinforcement learning for LLM post-training, yet its effectiveness in multilingual settings remains underexplored. We study OPD and its advanced variant, On-Policy Delta Distillation (OPD^2), for mathematical reasoning in English, Korean, and Japanese. OPD^2 improves OPD by using the probability gap between a post-trained teacher and its base model as the learning signal. Experiments with Qwen3 show that OPD^2 consistently outperforms the original OPD, with particularly strong improvements in Korean and Japanese, and generally narrows the English-Korean performance gap. We further find that English-only OPD can also increase performance for Korean and Japanese, but often shifts the responses toward English, highlighting the importance of multilingual data to preserving target-language responses.

📄 PDF Abstract BibTeX arXiv:2608.05802

Code (0)

등록된 구현이 없습니다.

Tasks

Mathematical ReasoningReinforcement Learning

Similar Papers 제목 키워드 기반

On-Policy Delta Distillation

2026-07-16 · Byeongho Heo, Jaehui Hwang, Sangdoo Yun, Dongyoon Han arxiv

On-policy distillation is an alternative post-training method in reinforcement learning that alleviates the constraints imposed by reward models by providing token-level supervision from a teacher model. Although on-poli…

Reinforcement Learning

RP-OPSD: Reasoning-Pivot-Guided On-Policy Self-Distillation for Multilingual Reasoning Transfer

2026-08-06 · Xinye Wang, Junxiao Liu, Shujian Huang arxiv

Multilingual reasoning transfer is crucial for extending reasoning capabilities of large language models (LLMs) beyond high-resource languages. On-policy self-distillation (OPSD) and its variants have emerged as a promis…

Mathematical ReasoningCross-Lingual Transfer

Crosslingual On-Policy Self-Distillation for Multilingual Reasoning

2026-05-10 · Yihong Liu, Raoyuan Zhao, Michael A. Hedderich, Hinrich Schütze arxiv

Large language models (LLMs) have achieved remarkable progress in mathematical reasoning, but this ability is not equally accessible across languages. Especially low-resource languages exhibit much lower reasoning perfor…

Reinforcement LearningMathematical Reasoning

DeltaPrompts: Escaping the Zero-Delta Trap in Multimodal Distillation

2026-05-15 · Jaehun Jung, Hyunwoo Kim, Brandon Cui, Ximing Lu 외 arxiv

Distillation enables compact Vision-Language Models (VLMs) to obtain strong reasoning capabilities, yet the prompts driving this process are typically chosen via simple heuristics or aggregated from off-the-shelf dataset…

Tool-Augmented Hybrid Ensemble Reasoning with Distillation for Bilingual Mathematical Problem Solving

2025-12-22 · Peiqing Lu, Yuan Zhang, Haoyun Zhang, Jiasen Zheng 외 arxiv

Bilingual mathematical problem solving needs a clear link between language reasoning and symbolic calculation. Large language models often handle language well but are weak in accurate computation. This paper presents HE…

Reinforcement LearningKnowledge DistillationMathematical Reasoning