PrentisAI releases GroundGUI-8B model for GUI agent tasks
PrentisAI introduced GroundGUI-8B, a multimodal model designed for precise GUI agent interactions. The model uses ScreenRef, a dataset of 586,000 screenshots and 633,000 referring expressions across desktop, web, and mobile environments. It combines supervised fine-tuning with Distillation-to-Reinforcement Learning (D2RL) to improve generalization and address sparse rewards in reinforcement…
Key points
- GroundGUI-8B model achieves 95.2% accuracy on ScreenSpot-V2 benchmark, surpassing 8B-scale competitors
- ScreenRef dataset contains 586K screenshots and 633K referring expressions for GUI grounding tasks
- D2RL method combines supervised fine-tuning with reinforcement learning for improved generalization
GroundGUI-8B outperforms 8B-scale models on benchmarks like ScreenSpot-V2 (95.2%) and MMBench-GUI L2 (86.8%), matching or exceeding performance of models trained on proprietary data. The project includes open-source code, models, and datasets, available at GitHub.
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- OpenAI offers U.S. university labs free ChatGPT Business seats with Rosalind access · 2 src
- Anthropic reports small AI models may self-replicate on IoT devices within months · 29 src
- Researchers propose LLM-referenced C-test for evaluating supra-intelligence · 8 src
- Researchers study permutation diffusion LLMs for recommendation reranking · 1 src
- Researchers compare Non-Generative verifier to Astra LLM for RAG claim checks · 1 src
Comments
via GitHub Discussions