An Elegant Academic Homepage Builder

Task-oriented AI Agent productivity platform

Task-oriented AI Agent productivity platform

Task-oriented AI Agent productivity platform

😼 优雅地使用基于 clash/mihomo 的代理环境

[ACL '26] This is the code repo for our ACL '26 Findings paper "MetaMem: Evolving Meta-Memory for Knowledge Utilization through Self-Reflective Symbolic Optimization"

[SIGIR '26] This is the code repo for our SIGIR’26 paper "ReAlign: Optimizing the Visual Document Retriever with Reasoning-Guided Fine-Grained Alignment".

Personal Homepage

你的赛博牛马斗兽场

[EACL] This is the code repo for our paper "Enhancing Knowledge Integration and Utilization of Large Language Models via Constructivist Cognition Modeling".

EdgeClaw: Edge-Cloud Collaborative Personal AI Assistant based on OpenClaw

[ACL '26] Lang2Act: Fine-Grained Visual Reasoning through Self-Emergent Linguistic Toolchains

Source code for our paper "Revealing the Attention Floating Mechanism in Masked Diffusion Models".

撸了么 - 今天你撸了吗？

Byted PyTorch Distributed for Hyperscale Training of LLMs and RLs

MinerU-HTML: An SLM-powered HTML main content extractor that outputs clean HTML bodies. Perfect for Deep Research Agents, RAG applications, and training data generation.

[ICASSP '26] This is the code repo for our paper: LegalΔ: Enhancing Legal Reasoning in LLMs via Reinforcement Learning with Chain-of-Thought Guided Information Gain.

[ADMA 2025 Best Paper Award] Code repo for our ADMA'25 paper: LegalDuet: Learning Fine-grained Representations for Legal Judgment Prediction via a Dual-View Contrastive Learning

Parsing-free RAG supported by VLMs

A Low-Code MCP Framework for Building Complex and Innovative RAG Pipelines

Official implementation of "VIRAL: Visual Representation Alignment for MLLMs".

verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework

UltraRAG 3.1

UltraRAG 2.0: Accelerating RAG for Scientific Research