About Me
- Hello, I'm Wei Liu (刘维). Welcome to my blog (0 views). You can search me on google with keyword "thinkwee", which means "The Thinking Wei".
- I'm interested in:
- (before 2023) Compression Intelligence in NLP.
- (2023 - 2025) Inference Time Scaling and Agentic AI.
- (2025 - now) AI in the Wild (check out this blog).
- Past Experience:
- 2014 - 2021: Bachelor of Communication Engineering in BUPT, and Master of Computer Engineering in CIST Lab@BUPT.
- 2021 - 2023: Application Research in the NLP&LLM Department in Tencent.
- 2023 - 2025: Working at THUNLP@Tsinghua University with Prof. Zhiyuan Liu and Prof. Chen Qian on LLM Multi-Agent System.
- 2025 - 2029(expected): Happy to join KCLNLP as a PhD candidate advised by Prof. Yulan He and Prof. Yali Du!
Recent News
- 2023.09.07 🚀 ChatDev made #1 on Github Trending! It now has earn over ... stars!
- 2024.05.15 🇹🇭 2 papers about LLM Multi-Agent System were accepted by ACL 2024!
- 2024.08.15 🚀 I made a interactive websites showing popular MultiAgent Frameworks and show THUNLP MultiAgent Team's amazing works at LLM MultiAgent EBook
- 2024.09.26 🇨🇦 1 paper about Personal Agentic AI was accepted by NeurIPS 2024!
- 2025.01.22 🇸🇬 1 paper about Scaling LLM Agents was accepted by ICLR 2025!
- 2025.05.15 🇦🇹 1 paper about LLM Agent for Software Development was accepted by ACL 2025!
- 2025.02.21 🇬🇧 I started my journey as a PhD student in the UK!
- 2025.05.16 🎉 Find KCLNLP's amazing works here, with 15 papers accepted by ACL 2025 and 3 papers accepted by ICML 2025!
- 2025.06.09 🚀 I made AgentsMeetRL, an awesome list of Reinforcement Learning-based Large Language Agent, which has now earned over ... stars! welcome to the era of experience!
- 2025.08.20 🎉 KCLNLP's new works here, with 8 papers accepted by EMNLP 2025 and 1 paper accepted by COLM 2025!
-
2025.08.20 🇨🇳
2 papers accepted by EMNLP 2025,
Check out
EcoLANG
and
NOVER!
- NOVER extends RLVR to any task, making a step towards the general reasoning model. See my detailed blog here.
- EcoLANG scales up the agentic social simulation faster and better with self-evolution on communication efficiency.
Services
- Program Committee Member
- ACL(2021,2022,2024)
- EMNLP(2020,2023,2024,2025)
- NeurIPS(2024,2025)
- ICLR(2024,2025,2026)
- CVPR(2025)
- AAAI(2025)
Industrial Experience
- I spent 3 years of happy time at Tencent, where I bring NLP in the
wild of Recommendation & Advertisement.
- Improving the NLU ability for News Feed Recommendation.
- Resolving the mismatch between commercial inclinations and content interests for Wechat Ads.
- Stability, Warm-Up, Efficiency-Quality Tradeoff, Interpretability & Explainability on Large Recommendation System.
- Diverse user interest modeling.
Publications
* denotes first/co-first author.
- Personal Agentic AI:
- Inference Time Scaling:
- Multi-Agents System with LLMs:
- EMNLP 2025 paper code EcoLANG: Efficient and Effective Agent Communication Language Induction for Social Simulation
- ACL 2024 paper code Communicative Agents for Software Development
- ACL 2024 paper code Experiential Co-Learning of Software-Developing Agents
- ACL 2025 paper code Multi-Agent Software Development through Cross-Team Collaboration
- arXiv 2024 paper code Iterative Experience Refinement of Software-Developing Agents
- Compression Intelligence in NLP:
- ACL 2021 paper code UniKeyphrase: A Unified Extraction and Generation Framework for Keyphrase Prediction*
- AAAI 2022 paper code Fast and Constrained Absent Keyphrase Generation by Prompt-Based Learning
- CoNLL 2021 paper code In Conclusion Not Repetition: Comprehensive Abstractive Summarization with Diversified Attention Based on Determinantal Point Processes*
- SIGIR 2019 paper CIST@CLSciSumm-19: Automatic Scientific Paper Summarization with Citances and Facets
- EMNLP 2020 paper CIST@CL-SciSumm 2020, LongSumm 2020: Automatic Scientific Document Summarization*
- RANLP 2019 paper code Multi-lingual Wikipedia Summarization and Title Generation On Low Resource Corpus*
- CCIS 2018 paper A Multi-View Abstractive Summarization Model Jointly Considering Semantics and Sentiment
- arXiv 2021 paper code CO2Sum: Contrastive Learning for Factual-Consistent Abstractive Summarization*
- arXiv 2021 paper code Subjective Bias in Abstractive Summarization*
Social Media
- I sometimes sync some articles from this blog on Wechat and Rednote, and I share other interesting things on these platforms. You can find me via above links.