8 papers
Controllable and Verifiable Tool-Use Data Synthesis for Agentic Reinforcement Learning
Siyuan Xu, Shiyang Li, Xin Liu +9
Existing synthetic tool-use corpora are primarily designed for offline supervised fine-tuning, yet reinforcement learning (RL) requires executable environments that support reward-…
Training LLMs for Multi-Step Tool Orchestration with Constrained Data Synthesis and Graduated Rewards
Cheng Jiayang, Xin Liu, Zhihan Zhang +8
Multi-step tool orchestration remains challenging for LLMs, as state-of-the-art models frequently fail on full sequence execution due to parameter errors. Training for these workfl…
Learning to Optimize Multi-Objective Alignment Through Dynamic Reward Weighting
Yining Lu, Zilong Wang, Shiyang Li +6
Prior work in multi-objective reinforcement learning typically uses linear reward scalarization with fixed weights, which provably fails to capture non-convex Pareto fronts and thu…
Self-Rewarding PPO: Aligning Large Language Models with Demonstrations Only
Qingru Zhang, Liang Qiu, Ilgee Hong +11
Supervised fine-tuning (SFT) has emerged as a crucial method for aligning large language models (LLMs) with human-annotated demonstrations. However, SFT, being an off-policy approa…
Train a Unified Multimodal Data Quality Classifier with Synthetic Data
Weizhi Wang, Rongmei Lin, Shiyang Li +7
The Multimodal Large Language Models (MLLMs) are continually pre-trained on a mixture of image-text caption data and interleaved document data, while the high-quality data filterin…
IHEval: Evaluating Language Models on Following the Instruction Hierarchy
Zhihan Zhang, Shiyang Li, Zixuan Zhang +11
The instruction hierarchy, which establishes a priority order from system messages to user messages, conversation history, and tool outputs, is essential for ensuring consistent an…