Chenxin Li

⭐

Chenxin Li | 李宸鑫

Hi! I'm Chenxin Li, a final-year Ph.D. candidate at The Chinese University of Hong Kong (CUHK). Before CUHK, I received my master's and bachelor's degrees from Xiamen University, along with a second bachelor's degree in Economics.

My recent interest lies in scaling LLM/VLM agents for digital automation, including: (i) Coding agents for general computer use (Claw-Eval-Live, Seed Auto R&D Coding Agent, BlenderAgent, LightroomPSAgent, On-Policy Data Evolution, Hybrid CLI-GUI Harness) and (ii) Visual coding agents for design artifacts (IRBlender, JarvisArt, JarvisIR, Seed code-to-chart/web). These experiences span building frontier agent scaffolds/harnesses, task/verifier/benchmark, trajectory interaction/distillation, and mid/post-training for agents.

Throughout my fulfilling years in Master & Ph.D., I have gained extensive industry exposure through internships at ByteDance Seed, Tencent AI, Ant Ling, Giga AI, AMD, Hedra AI, JoinQuant, etc. These experiences taught me to stay open to cross-disciplinary opportunities across agents, world models, and quantitative research. More broadly, I build in an AI-native way, using hands-on artifacts to learn agent boundaries and share explorations through 📝 Blogs and 🛠️ OpenAgentLabs.

I anticipate graduating in the summer of 2026 and am interested in industrial positions (Profile). Please feel free to reach out via email (chenxinli@link.cuhk.edu.hk) or WeChat (jasonchenxinli).

🚀 Selected Work

* Equal contribution, † Project Leader, ‡ Corresponding author

Claw-Eval-Live: A Live Agent Benchmark for Evolving Real-World Workflows

Chenxin Li†, Zhengyang Tang, Huangxin Lin, Yunlong Lin, Shijue Huang, Shengyuan Liu, Bowen Ye, Rang Li, Lei Li, Benyou Wang, Yixuan Yuan

[Project] [Paper] [Code]

A live workflow-agent benchmark with refreshable demand signals and verifiable execution traces; 105 tasks across 22 categories, 13 frontier models, top model passes only 66.7%.

Towards On-Policy Data Evolution for Visual-Native Multimodal Deep Search Agents

Shijue Huang, Hangyu Guo, Chenxin Li, Junting Lu, Xinyu Geng, Zhaochen Su, Zhenyu Li, Shuang Chen, Hongru Wang, Yi R Fung

[Project] [Paper] [Code]

An on-policy data-evolution framework for visual-native multimodal deep-search agents, using agent rollouts to evolve training data and improve search behavior.

HybridHarness: A Mixed-Action Orchestration Harness and Benchmark for Phone Agents across CLI, GUI, and MCP Tools

[Project] [Paper] [Code] [Dataset]

A mixed-action phone-agent harness and benchmark that routes across CLI, GUI, and MCP tools with trace-backed verification.

Seed-1.8: Towards Generalized Real-World Agency

ByteDance Seed Team

[Project] [Model Card]

Contributed to agent post-training for visual coding and agentic tool-use.

UI-TARS-2: Advancing GUI Agent with Multi-Turn Reinforcement Learning

ByteDance Seed Team

[Project] [Report]

Contributed to agent post-training.

Ling: Open-sourced LLM with MoE Architecture by InclusionAI

Ant Group InclusionAI Team

[Project]

Contributed to agentic memory.

🛠️ Open Agent Labs

I build agent tools that help people work smarter by turning repetitive, inefficient workflows into AI-native experiences.

Agent Context Manager Keep coding-agent work coherent across sessions, tools, and context resets. Context · Agent Memory · Harness

Open Codex Computer Use Bring computer-use into general agents like Codex, letting them operate your computer in parallel without interrupting you. MCP · Computer Use · macOS

OpenReview Agent Reduce brittle manual steps in OpenReview submission and venue-transfer workflows. OpenReview · Agent Skill · Submission

AI-native Slides Turn rough ideas into polished, editable decks with agentic story shaping and QA. Slides · Agent Skill · Presentation Workflow

IRBlenderAgent: Evaluating Vision-Language Model Scene Understanding as Agentic Inverse Rendering

Parker Liu*, Chenxin Li*, Zhengxin Li, Yipeng Wu, Wuyang Li, Zhiqin Yang, Zhenyuan Zhang, Yunlong Lin, Sirui Han, Brandon Y. Feng

NeurIPS 2025

[Project] [Paper] [Code]

An agentic inverse-rendering framework that closes the loop from visual understanding to structured code generation, Blender execution, and environment feedback.

U-KAN Makes Strong Backbone for Visual Understanding and Generation

Chenxin Li*, Xinyu Liu*, Wuyang Li*, Cheng Wang*, Hengyu Liu, Yixuan Yuan

AAAI 2025

[Project] [Paper] [Code] 🏆 2025 PaperDigest Most Influential Paper

Integrating Linear Attention mechanism like KAN into vision backbone

JarvisArt: Liberating Human Artistic Creativity via an Intelligent Photo Retouching Agent

Yunlong Lin*, Zixu Lin*, Kunjie Lin*, Chenxin Li*, Haoyu Chen, Zhongdao Wang, Xinghao Ding†, Wenbo Li, Shuicheng Yan†

NeurIPS 2025

[Project] [Paper] [Code]

An artistic editing agent system that plans multi-step retouching commands and coordinates expert models for execution-quality image refinement.

Visual Large Language Model Fine-Tuning via Simple Parameter-Efficient Modification

Mengjiao Li, Zhiyuan Ji, Chenxin Li†, Lianliang Nie, Zhiyang Li, Masashi Sugiyama

EMNLP 2024

[Project] [Paper] [Code]

Simple yet efficient parameter-efficient fine-tuning for VLM alignment.

🧭 Selected Experience

LLM Agent

ByteDance Seed: Agent for CLI, visual coding and GUI (Seed-1.6 / Seed-1.8 / UI-TARS-2)
Tencent AI Lab: Agent execution loop, code generation and environment-grounded RL (IRBlender-Bench)
Ant Ling: Agent memory, context compression and output verification (Ling-Pilot)
AMD: Visual token compression for efficient VLMs

World Model

Giga AI: World-model agent for 3D environments and tool-augmented training (Giga Brain-0)
Hedra AI: Omnimodal attention architecture for commercial digital-human generation (Hedra Character-3)

Quant

JoinQuant: RL post-training of LLMs for quant alpha-factor mining; agent evaluation on financial data

ScholaGO (Co-founder): LLM4Education Startup

Co-founded ScholaGO Education Technology Company Limited (学旅通教育科技有限公司) to build LLM-powered education products that turn static content into immersive, interactive, multimodal learning experiences. Grateful to receiving funding from HKSTP, HK Tech 300, and Alibaba Cloud.

🎓 Professional Activities

Workshop Organizer: AIM-FM: Advancements In Foundation Models Towards Intelligent Agents (NeurIPS 2024)
Talks: "UKAN" at VALSE Summit (Jun 2025) and DAMTP, University of Cambridge (Jul 2024)
Conference Reviewer: ICLR, NeurIPS, ICML, ACL, CVPR, ICCV, ECCV, EMNLP, AAAI, ACM MM
Journal Reviewer: Nature Machine Intelligence, PAMI, TIP, DMLR, PR, TNNLS

🌱 Hobbies Beyond Work

Crafting Agent Artifacts: I enjoy building agent tools hands-on because the rapid feedback loop of vibe coding gives me a concrete feel for where different models excel, fail, and reach their limits. Over time, this has shaped an increasingly AI-native way of thinking.
Reading: I read history, philosophy, and sociology over the long term. I enjoy reasoning from first principles to anticipate future trends and make my bets accordingly.
Stock Investment: I see investing as real-world RL: making decisions, receiving feedback, and refining strategies. To me, many choices are bets; as agents dramatically amplify productivity, what matters shifts to where we "invest" our attention.