11 papers
ANDES: Agent Native Data Evolving Synthesis Tool for Autonomous Instruction Alignment
Zhengyang Zhao, Shengjie Ye, Lu Ma +3
AI agents are increasingly being tasked with automating AI research itself, particularly the critical post-training phase that transforms base LLMs into aligned assistants. However…
Training with Harnesses: On-Policy Harness Self-Distillation for Complex Reasoning
Zhengyang Zhao, Lu Ma, Wentao Zhang
Inference-time harnesses substantially improve large language models on complex reasoning tasks. However, the intrinsic capabilities of the underlying model remain unchanged by the…
DataFlex: A Unified Framework for Data-Centric Dynamic Training of Large Language Models
Hao Liang, Zhengyang Zhao, Meiyi Qiang +22
Data-centric training has emerged as a promising direction for improving large language models (LLMs) by optimizing not only model parameters but also the selection, composition, a…
GIFT: Reconciling Post-Training Objectives via Finite-Temperature Gibbs Initialization
Zhengyang Zhao, Lu Ma, Yizhen Jiang +7
The prevailing post-training paradigm for Large Reasoning Models (LRMs) - Supervised Fine-Tuning (SFT) followed by Reinforcement Learning (RL) - suffers from an intrinsic optimizat…
Learning What Reinforcement Learning Can't: Interleaved Online Fine-Tuning for Hardest Questions
Lu Ma, Hao Liang, Meiyi Qiang +9
Recent advances in large language model (LLM) reasoning have shown that sophisticated behaviors such as planning and self-reflection can emerge through reinforcement learning (RL).…
Thinking by Subtraction: Confidence-Driven Contrastive Decoding for LLM Reasoning
Lexiang Tang, Weihao Gao, Bingchen Zhao +4
Recent work on test-time scaling for large language model (LLM) reasoning typically assumes that allocating more inference-time computation uniformly improves correctness. However,…