Improve model card and link to paper

README.md

@@ -8,30 +8,42 @@ datasets:
 - MVBench
 - TempCompass
 - Video-MME
-language: en
+language:
+- en
+library_name: transformers
 license: mit
+pipeline_tag: video-text-to-text
 tags:
 - video-understanding
 - reasoning
 - multimodal
 - reinforcement-learning
 - question-answering
-library_name: transformers
-pipeline_tag: video-text-to-text
 ---
 
-# Paper abstract
+# VideoThinker-R1-3B
+
+[**Beyond Perceptual Shortcuts: Causal-Inspired Debiasing Optimization for Generalizable Video Reasoning in Lightweight MLLMs**](https://huggingface.co/papers/2605.01324)
+
+VideoThinker is a causal-inspired framework that enables lightweight multimodal language models (3B parameters) to achieve robust video reasoning. It addresses the phenomenon of "perceptual bias," where reinforcement learning can compel lightweight models to adopt perceptual shortcuts from data rather than developing genuine reasoning abilities.
 
-The abstract of the paper is the following:
+The framework employs a two-stage debiasing process:
+1. **Bias Aware Training**: Forges a dedicated "bias model" to embody shortcut behaviors.
+2. **Causal Debiasing Policy Optimization (CDPO)**: Fine-tunes the primary model using a repulsive objective to push it away from the bias model's flawed logic.
 
-Although reinforcement learning (RL) has significantly advanced reasoning capabilities in large multimodal language models (MLLMs), its efficacy remains limited for lightweight models essential for edge deployments.To address this issue, we leverage causal analysis and experiment to reveal the underlying phenomenon of perceptual bias, demonstrating that RL-based fine-tuning compels lightweight models to preferentially adopt perceptual shortcuts induced by data biases, rather than developing genuine reasoning abilities.Motivated by this insight, we propose VideoThinker, a causal-inspired framework that cultivates robust reasoning in lightweight models through a two-stage debiasing process. First, the Bias Aware Training stage forges a dedicated "bias model" to embody these shortcut behaviors. Then, the Causal Debiasing Policy Optimization (CDPO) algorithm fine-tunes the primary model, employing an innovative repulsive objective to actively push it away from the bias model's flawed logic while simultaneously pulling it toward correct, generalizable solutions.Our model, VideoThinker-R1, establishes a new state-of-the-art in video reasoning efficiency. For same-scale comparison, requiring no Supervised Fine-Tuning (SFT) and using only 1 of the training data for RL, it surpasses VideoRFT-3B with a 3.2% average gain on widely-used benchmarks and a 7% lead on VideoMME. For cross-scale comparison, it outperforms the larger Video-UTR-7B model on multiple benchmarks, including a 2.1% gain on MVBench and a 3.8% gain on TempCompass.
+## Performance
+VideoThinker-R1 establishes a new state-of-the-art in video reasoning efficiency. Using only 1K training samples and no Supervised Fine-Tuning (SFT), it:
+- Surpasses VideoRFT-3B by 7% on VideoMME.
+- Outperforms larger models (e.g., Video-UTR-7B) on reasoning-heavy benchmarks like MVBench and TempCompass.
 
-This repository contains the model as presented in "Beyond Perceptual Shortcuts: Causal-Inspired Debiasing Optimization for Generalizable Video Reasoning in Lightweight MLLMs".
+## Resources
+- **Code**: [GitHub - falonss703/VideoThinker](https://github.com/falonss703/VideoThinker)
+- **Paper**: [Hugging Face Papers](https://huggingface.co/papers/2605.01324)
 
-For training and evaluation, please refer to the Code: https://github.com/falonss703/VideoThinker
+## Citation
+If you find this project useful in your research, please consider citing:
 
-If you find this project useful in your research, please consider cite:
-```BibTeX
+```bibtex
 @inproceedings{wu2026videothinker,
   title={Beyond Perceptual Shortcuts: Causal-Inspired Debiasing Optimization for Generalizable Video Reasoning in Lightweight MLLMs},
   author={Wu, Jingze and Zhang, Quan and Suo, Hongfei and Cai, Zeqiang and Chen, Hongbo},