Skip to content
@TIGER-AI-Lab

TIGER Lab

Our lab is currently based in UWaterloo, focusing on Text and Image Generative Research

Text and Image GEnerative Research (TIGER) Lab

In Chinese, our lab is named "虎头帮".

This is the official website for TIGER AI Lab (https://wenhuchen.github.io/lab).

About Us

Welcome to the Text and Image GEnerative Research (TIGER) Lab at the University of Waterloo. Our lab is dedicated to enabling more controllable and faithful generation of both text and images.

Our Goal

Our primary goal is to advance GenAI (Generative Artificial Intelligence) and make it more applicable to society. We believe in the transformative potential of AI and strive to develop innovative solutions that can revolutionize the way we create digitalized content. Our research focuses on a variety of topics within the realm of GenAI, including but not limited to:

  • Text generation
  • Image and Video Generation
  • Multi-modal Retrieval and Grounding
  • Improving Reasoning and Planning in GenAI
  • Controllability in GenAI
  • Evaluation in GenAI

Our Team

We are a dynamic team of dedicated researchers and faculty members. Our team is constantly growing and we welcome individuals who share our passion for advancing GenAI.

Publications

Our work has been published in several esteemed journals and conferences. You can find a list of our publications here).

Huggingface

Our dataset, checkpoints and demo are all being released under TIGER-Lab.

Pinned Loading

  1. .github .github Public

    main page

Repositories

Showing 10 of 64 repositories
  • ClawBench Public

    Open-source benchmark for browser AI agents on daily tasks.

    TIGER-AI-Lab/ClawBench's past year of commit activity
    Python 622 Apache-2.0 45 45 11 Updated Aug 31, 2026
  • RewardHarness Public

    Self-evolving agentic reward framework for image-editing evaluation — 47.4% on EditReward-Bench from only 100 preference demos, no reward-model training. arXiv 2605.08703.

    TIGER-AI-Lab/RewardHarness's past year of commit activity
    Python 34 8 2 0 Updated Aug 25, 2026
  • VLM2Vec Public

    This repo contains the code for "VLM2Vec / MMEB" [ICLR 2025], "VLM2Vec-V2 / MMEB-V2" [TMLR 2026], and "MMEB-V3" [COLM 2026]

    TIGER-AI-Lab/VLM2Vec's past year of commit activity
    Python 680 Apache-2.0 64 29 0 Updated Aug 23, 2026
  • Context-Forcing Public

    Context Forcing: Consistent Autoregressive Video Generation with Long Context [ICML26]

    TIGER-AI-Lab/Context-Forcing's past year of commit activity
    Python 111 2 1 0 Updated Aug 4, 2026
  • Pixel-Reasoner Public

    Pixel-Level Reasoning Model trained with RL [NeuIPS25]

    TIGER-AI-Lab/Pixel-Reasoner's past year of commit activity
    Python 307 MIT 13 5 0 Updated Jul 28, 2026
  • StructEval Public

    Evaluating LLMs' abilities to generate structural output [TMLR2025]

    TIGER-AI-Lab/StructEval's past year of commit activity
    Python 23 Apache-2.0 7 2 4 Updated Jul 26, 2026
  • EditReward Public

    EditReward: A Human-Aligned Reward Model for Instruction-Guided Image Editing [ICLR 2026]

    TIGER-AI-Lab/EditReward's past year of commit activity
    Python 159 MIT 6 4 0 Updated Jul 26, 2026
  • verl-tool Public

    A version of verl to support diverse tool use [TMLR 2026]

    TIGER-AI-Lab/verl-tool's past year of commit activity
    Python 1,037 MIT 89 8 1 Updated Jul 15, 2026
  • TIGER-AI-Lab/FIM-Midtraining's past year of commit activity
    Python 20 Apache-2.0 1 1 0 Updated Jul 15, 2026
  • SWE-QA-Pro Public

    SWE-QA-Pro: A Representative Benchmark and Scalable Training Recipe for Repository-Level Code Understanding [ACL 2026]

    TIGER-AI-Lab/SWE-QA-Pro's past year of commit activity
    Python 13 MIT 2 0 0 Updated Jun 26, 2026