【本日の論文】
1. SynthDocBench: Controlled Benchmark for Long-Context Visual Document Understanding
https://huggingface.co/papers/2607.10400
2. Read It Back: Pretrained MLLMs Are Zero-Shot Reward Models for Text-to-Image Generation
https://huggingface.co/papers/2607.11886
3. Search Beyond What Can Be Taught: Evolving the Knowledge Boundary in Agentic Visual Generation
https://huggingface.co/papers/2607.05382
4. Blind-Spots-Bench: Evaluating Blind Spots in Multimodal Models
https://huggingface.co/papers/2607.08317
5. Function-Aware Fill-in-the-Middle as Mid-Training for Coding Agent Foundation Models
https://huggingface.co/papers/2607.12463