Thinkwee's Blog

Too Stupid to Give Up Learning

About Me

I'm in London

Since February 2025, I've been based in London pursuing my PhD journey!
Fun Fact: My giraffe icon isn't random - I actually have a long neck!
See for yourself and meet other talented researchers on our KCLNLP team page.

You can find this guy via
Linkedin Google Scholar GitHub Gmail

wechat

About Me

I'm in BUPT
I'm in Tecent
I'm in Tsinghua
  • Hello, I’m Wei Liu (刘维). Welcome to my blog (0 views</span>). You can search me on google with keyword “thinkwee”, which means “The Thinking Wei”.
  • 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-now: Proud to be a PhD advised by Prof. Yulan He and a member of KCLNLP!

Recent News

  • 2025.5.16 Checkout KCLNLP’s amazing works here, with 15 papers accepted by ACL 2025 and 3 papers accepted by ICML 2025!.
  • 2025.5.21 Check out NOVER, a novel verifier-free reinforcement learning framework for training Large Reasoning Model. Train your own R1-Zero-like reasoning model on ANY DATA!
  • 2025.6.9 Check out AgentsMeetRL, an awesome list of Reinforcement Learning-based Large Language Agent!

Research Interests

  • Inference Time Scaling and Agentic AI.
  • Compression Intelligence in NLP.
  • Served as reviewer for
    • ACL(2021,2022,2024)
    • EMNLP(2020,2023,2024,2025)
    • NeurIPS(2024,2025)
    • ICLR(2024)
    • CVPR(2025)

Industrial Experience

  • At Tencent, I aim to bridge the gap between technology in NLP and scenario in 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:
    • NeurIPS 2024 paper code Autonomous Agents for Collaborative Task under Information Asymmetry*
  • Inference Time Scaling:
    • arXiv 2025 paper code NOVER: Incentive Training for Language Models via Verifier-Free Reinforcement Learning*
    • ICLR 2025 paper code Scaling Large-Language-Model-based Multi-Agent Collaboration
  • Multi-Agents System with LLMs:
    • 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
    • arXiv 2025 paper code EcoLANG: Efficient and Effective Agent Communication Language Induction for Social Simulation
  • 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*
0%