TIGER-AI-Lab/Pixel-Reasoner
Pixel-Level Reasoning Model trained with RL [NeuIPS25]
This project helps advanced AI researchers and practitioners enable Vision-Language Models (VLMs) to "think" more deeply about images and videos. By teaching VLMs to perform operations like zooming in or selecting specific frames, it allows them to analyze visual evidence directly. The output is a VLM that can reason more effectively in complex visual tasks.
282 stars.
Use this if you are a researcher or AI engineer working on advanced VLM capabilities and need to train models that perform detailed visual reasoning beyond simple captioning or classification.
Not ideal if you are looking for an out-of-the-box solution for basic image analysis or text generation, or if you don't have experience with large model training and reinforcement learning.
Stars
282
Forks
11
Language
Python
License
MIT
Category
Last pushed
Nov 06, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/transformers/TIGER-AI-Lab/Pixel-Reasoner"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
cvs-health/uqlm
UQLM: Uncertainty Quantification for Language Models, is a Python package for UQ-based LLM...
PRIME-RL/TTRL
[NeurIPS 2025] TTRL: Test-Time Reinforcement Learning
sapientinc/HRM
Hierarchical Reasoning Model Official Release
tigerchen52/query_level_uncertainty
query-level uncertainty in LLMs
reasoning-survey/Awesome-Reasoning-Foundation-Models
✨✨Latest Papers and Benchmarks in Reasoning with Foundation Models