2 papers
cs.LG2026
Bridging SFT and RL: Dynamic Policy Optimization for Robust Reasoning
Taojie Zhu, Dongyang Xu, Ding Zou +4
Post-training paradigms for Large Language Models (LLMs), primarily Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL), face a fundamental dilemma: SFT provides stability…
cs.CV2025
Revisiting the Data Sampling in Multimodal Post-training from a Difficulty-Distinguish View
Jianyu Qi, Ding Zou, Wenrui Yan +5
Recent advances in Multimodal Large Language Models (MLLMs) have spurred significant progress in Chain-of-Thought (CoT) reasoning. Building on the success of Deepseek-R1, researche…