2026
4 updates Mar 20
🚀 Released Nemotron-Cascade 2, a compact open 30B MoE with Gold Medal-level performance on the IMO, IOI, and ICPC World Finals, while using 20x fewer parameters than frontier open models. Model and data are available on Hugging Face.
Mar 11
🚀 Released Nemotron 3 Super, an open and efficient hybrid Mamba-Transformer MoE built for strong agentic reasoning.
Feb 9
🚀 Released OpenResearcher, a fully open pipeline for synthesizing long-horizon deep research trajectories with open data, models, and demo.
Jan 26
🎉 Critique-Coder was accepted to ICLR 2026.
2025
11 updates Dec 27
🎉 QuickVideo was accepted to TMLR.
Dec 12
🎉 StructEval was accepted to TMLR, received the Journal-to-Conference Certificate, and will be presented at ICLR 2026.
Dec 1
🏆 Mantis received the TMLR 2025 Outstanding Paper Award.
Sep 26
🎬 Released VideoScore2, a multi-dimensional and interpretable evaluator for generative videos with detailed reasoning traces.
Sep 1
📄 Released the VerlTool technical report, highlighting a unified ARLT framework with async execution, multi-tool support, and competitive results across 6 domains.
Aug 18
💼 Started my internship at NVIDIA ADLR.
Jun 1
🚀 Released the VerlTool codebase, a unified framework for training tool-using language agents with async rollouts and modular tool APIs.
May 20
🚀 Released General-Reasoner, extending RL-style reasoning beyond math and code with verified web-scale data and a generative answer verifier.
Mar 8
Will join NVIDIA as a research intern in Santa Clara this summer.
Feb 3
Released AceCoder and its SoTA reward model for coding.
Jan 22
MEGA-Bench was accepted to ICLR 2025.
2024
14 updates Nov 14
Sep 26
Sep 19
VideoScore was accepted to EMNLP 2024.
Jun 24
Started my internship at 
today!
today! Jun 23
Jun 18
I arrived in Seattle for the CVPR 2024 conference.
May 10
🐯TIGERScore is accepted to TMLR 2024!
May 3
Apr 14
We release 
Mantis, enhancing LMM with Interleaved Multi-Image Instruction Tuning!
Mantis, enhancing LMM with Interleaved Multi-Image Instruction Tuning! Apr 8
WildVision Arena has been accepted to CVPR 2024 demo track and will be presented at the conference!
Apr 5
MMMU is accepted to CVPR 2024 oral presentation! Feb 20
I am excited to announce that I will be joining AI2 Mosaic Team as a research intern this summer!
Feb 12
GenAI arena is now also online! You can test popular image generation/editing models here!
Feb 7
Check out our WildVision Arena demo on HuggingFace for VLMs!
2023
5 updates Dec 1
We release PairRM-0.4B 🤗 based on LLM-Blender!
Nov 29
We release the new benchmark MMMU for evaluating multi-modal models!
MMMU for evaluating multi-modal models! Oct 4
Check my first work at UW: 🐯TIGERScore!
Sep 2
I arrived at University of Waterloo and started my Ph.D. journey!
Jun 5
We release LLM-Blender! It’s accepted to ACL 2023!