Official Implementation of GenRxR (RecSys'26) (Paper: https://arxiv.org/abs/2607.24829)

The source code of RES-MR

OpenCoF: Learning to Reason Through Video Generation

a VERL training framework that decouple reasoning and confidence in rewards

Implementation of KDD'26 paper "Embedding-Space Orthogonal Decomposition for Robust Social Recommendation".

[NeurIPS 2025] 3D-GSRD: 3D Molecular Graph Auto-Encoder with Selective Re-mask Decoding

Official Implementation for Paper "Demystifying Multi-Agent Debate: The Role of Confidence and Diversity"

【ICML2026 Spotlight】 T2PO: Uncertainty-Guided Exploration Control for Stable Multi-Turn Agentic Reinforcement Learning

Dr. MAS is an end-to-end RL training framework for multi-agent LLM systems, supporting the co-training of multiple (heterogeneous) LLMs.

An extremely modular, easy-to-use, and research-oriented framework for Generative Recommendation.

An awesome list of papers on trustworthy LLM-empowered recommendation

Official implementation of "TROJail: Trajectory-Level Optimization for Multi-Turn Large Language Model Jailbreaks with Process Rewards"

[KDD'26] Uncertainty-aware Generative Recommendation

Claude Code to OpenAI API Proxy

[ICML'24 Spotlight] "TravelPlanner: A Benchmark for Real-World Planning with Language Agents"

We propose the first Tri-party LLM-agent Recommendation framework (TriRec) that explicitly coordinates user utility, item exposure, and platform-level fairness.

[ICDE'24] Code of "Adapting Large Language Models by Integrating Collaborative Semantics for Recommendation."

[ACL 2025] iAgent: LLM Agent as a Shield between User and Recommender Systems

[NeurIPS 2025] MINT-CoT: Enabling Interleaved Visual Tokens in Mathematical Chain-of-Thought Reasoning

[CVPR2025] Precise, Fast, and Low-cost Concept Erasure in Value Space: Orthogonal Complement Matters