Daily Papers of 2025-07-31

1. ScreenCoder: Advancing Visual-to-Code Generation for Front-End Automation via Modular Multimodal Agents 86 upvotes, #1 of 2025-07-31
2. BANG: Dividing 3D Assets via Generative Exploded Dynamics 60 upvotes, #2 of 2025-07-31
3. Falcon-H1: A Family of Hybrid-Head Language Models Redefining Efficiency and Performance 60 upvotes, #2 of 2025-07-31
4. VL-Cogito: Progressive Curriculum Reinforcement Learning for Advanced Multimodal Reasoning 41 upvotes, #4 of 2025-07-31
5. MetaCLIP 2: A Worldwide Scaling Recipe 22 upvotes, #5 of 2025-07-31
6. Step-3 is Large yet Affordable: Model-system Co-design for Cost-effective Decoding 17 upvotes, #6 of 2025-07-31
7. Adapting Vehicle Detectors for Aerial Imagery to Unseen Domains with Weak Supervision 10 upvotes, #7 of 2025-07-31
8. MixGRPO: Unlocking Flow-based GRPO Efficiency with Mixed ODE-SDE 10 upvotes, #7 of 2025-07-31
9. Towards Omnimodal Expressions and Reasoning in Referring Audio-Visual Segmentation 9 upvotes, #9 of 2025-07-31
10. Efficient Differentially Private Fine-Tuning of LLMs via Reinforcement Learning 8 upvotes, #10 of 2025-07-31
11. Repair-R1: Better Test Before Repair 8 upvotes, #10 of 2025-07-31
12. DreamScene: 3D Gaussian-based End-to-end Text-to-3D Scene Generation 6 upvotes, #12 of 2025-07-31

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.