Towards On-Policy SFT: Distribution Discriminant Theory and its Applications in LLM Training

• URL: http://arxiv.org/abs/2602.12222v1
• Date: Thu, 12 Feb 2026 17:59:58 GMT
• Title: Towards On-Policy SFT: Distribution Discriminant Theory and its Applications in LLM Training
• Authors: Miaosen Zhang, Yishan Liu, Shuxia Lin, Xu Yang, Qi Dai, Chong Luo, Weihao Jiang, Peng Hou, Anxiang Zeng, Xin Geng, Baining Guo,
• Abstract summary: Supervised fine-tuning (SFT) is computationally efficient but often yields inferior generalization compared to reinforcement learning (RL)<n>We propose a framework to bridge this chasm by enabling On-Policy SFT.
• Score: 61.1421888242439
• License: http://creativecommons.org/publicdomain/zero/1.0/
• Abstract: Supervised fine-tuning (SFT) is computationally efficient but often yields inferior generalization compared to reinforcement learning (RL). This gap is primarily driven by RL's use of on-policy data. We propose a framework to bridge this chasm by enabling On-Policy SFT. We first present \textbf{\textit{Distribution Discriminant Theory (DDT)}}, which explains and quantifies the alignment between data and the model-induced distribution. Leveraging DDT, we introduce two complementary techniques: (i) \textbf{\textit{In-Distribution Finetuning (IDFT)}}, a loss-level method to enhance generalization ability of SFT, and (ii) \textbf{\textit{Hinted Decoding}}, a data-level technique that can re-align the training corpus to the model's distribution. Extensive experiments demonstrate that our framework achieves generalization performance on par with prominent offline RL algorithms, including DPO and SimPO, while maintaining the efficiency of an SFT pipeline. The proposed framework thus offers a practical alternative in domains where RL is infeasible. We open-source the code here: https://github.com/zhangmiaosen2000/Towards-On-Policy-SFT

Related papers

• Trust-Region Adaptive Policy Optimization [82.09255251747818]
  Post-training methods play an important role in improving large language models' (LLMs) complex reasoning abilities.<n>We introduce TRAPO, a framework that interleavesSupervised Fine-Tuning (SFT) and Reinforcement Learning (RL) within each training instance.<n>Experiments on five mathematical reasoning benchmarks show that TRAPO consistently surpasses standard SFT, RL, and SFT-then-RL pipelines.
  arXiv  Detail & Related papers  (2025-12-19T14:37:07Z)
• Self-Rewarding PPO: Aligning Large Language Models with Demonstrations Only [70.43369087819332]
  Supervised fine-tuning (SFT) has emerged as a crucial method for aligning large language models with human-annotated demonstrations.<n>We propose Self-Rewarding PPO, a novel fine-tuning method that leverages on-policy techniques to enhance generalization performance.
  arXiv  Detail & Related papers  (2025-10-24T02:02:13Z)
• UniAPL: A Unified Adversarial Preference Learning Framework for Instruct-Following [12.924923059340395]
  We argue that post-training alignment is fundamentally a unified Preference Learning problem.<n>UniAPL implements a single-stage unified training objective, jointly learning from mixed batches of SFT and preference data.
  arXiv  Detail & Related papers  (2025-09-29T17:53:09Z)
• Group-Relative REINFORCE Is Secretly an Off-Policy Algorithm: Demystifying Some Myths About GRPO and Its Friends [64.71326476563213]
  Off-policy reinforcement learning for large language models (LLMs) is attracting growing interest.<n>We present a first-principles derivation for grouprelative REINFORCE without assuming a specific training data distribution.<n>This perspective yields two general principles for adapting REINFORCE to off-policy settings.
  arXiv  Detail & Related papers  (2025-09-29T02:34:54Z)
• AMFT: Aligning LLM Reasoners by Meta-Learning the Optimal Imitation-Exploration Balance [7.685078284407324]
  Large Language Models (LLMs) are typically fine-tuned for reasoning tasks through a two-stage pipeline Supervised Fine-Tuning (SFT) followed by Reinforcement Learning (RL)<n>Recent single-stage methods attempt to unify SFT and RL using principleds, but lack a mechanism for dynamically balancing the two paradigms.<n>We introduce textbf Meta Fine-Tuning (AMFT), a novel single-stage algorithm that learns the optimal balance between SFT's implicit, path-level reward and RL's explicit, outcome-based reward.
  arXiv  Detail & Related papers  (2025-08-09T11:40:54Z)
• Implicit Reward as the Bridge: A Unified View of SFT and DPO Connections [65.36449542323277]
  We present a unified theoretical framework bridgingSupervised Fine-Tuning (SFT) and preference learning in Large Language Model (LLM) post-training.<n>We propose a simple yet effective learning rate reduction approach that yields significant performance improvements.
  arXiv  Detail & Related papers  (2025-06-15T05:42:29Z)
• DUMP: Automated Distribution-Level Curriculum Learning for RL-based LLM Post-training [25.91869315787235]
  We present a principled curriculum learning framework grounded in the notion of distribution-level learnability.<n>Our framework prioritizes distributions with either high average advantage (exploitation) or low sample count (exploration)<n>Our experiments show that our framework significantly improves convergence speed and final performance.
  arXiv  Detail & Related papers  (2025-04-13T20:10:27Z)
• Towards Deployment-Efficient Reinforcement Learning: Lower Bound and Optimality [141.89413461337324]
  Deployment efficiency is an important criterion for many real-world applications of reinforcement learning (RL) We propose a theoretical formulation for deployment-efficient RL (DE-RL) from an "optimization with constraints" perspective.
  arXiv  Detail & Related papers  (2022-02-14T01:31:46Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.