University of Science and Technology of China
We track 9 benchmarks that University of Science and Technology of China made or helped make.
0 in use 6 with public code 4 with a leaderboard here
Benchmarks
| Name | Kind | Released | Best score | Status |
|---|---|---|---|---|
| DeepResearch Bench Agents write cited research reports for 100 PhD-level tasks across 22 fields. Judges score report quality against references and check citation accuracy. | Deep research | Jun 2025 | 55.3% | Fading |
| PhysVista Vision-language models perceive, reason about and judge whether videos follow real-world physics. The benchmark also tests if they can judge the physical realism of generated videos. | Physical prediction | Sep 2026 | 85% | Not in cards |
| AsyncTool Agents coordinate several tool-use tasks while tool responses arrive after simulated delays. The evaluator checks function calls, parameters, task completion, and the final environment state. Scheduling metrics measure whether agents use waiting time effectively. | Function calling | May 2026 | 38.1% | Not in cards |
| TextVBench Models render requested text in video or preserve text from a reference image. The score tracks spelling, placement, and consistency across frames. | Video generation | Oct 2026 | | Not in cards |
| UFO-Bench Models generate customized images from text and visual references. The evaluator breaks each request into individual requirements and checks how they interact. | Image generation and editing | Sep 2026 | | Not in cards |
| DeepResearch Bench II Deep research agents write reports for 132 research briefs in 22 domains, each derived from a review article they may not consult. A judge model checks 9,287 expert criteria. | Deep research | Jan 2026 | 43.1% | Not in cards |
| DeepResearch Bench II (Gemini 2.5 Pro judge) Systems conduct deep research under the legacy Gemini 2.5 Pro evaluator. Scores remain separate from the current GPT-5.5 evaluator. | Deep research | Jan 2026 | 45.4% | Not in cards |
| RoboCasa-Mem Robots complete nine household tasks that require memory. Later steps depend on objects the robot saw or actions it took earlier. | Simulated robots | Nov 2026 | 16% | Not in cards |
| ExeWorldBench Coding agents build executable code-based 3D worlds that capture how a world works, testing world understanding and its realization in code. | Physical prediction | Oct 2026 | Not in cards |