HLL Benchmark
Explore score history ↗HLL is a benchmark that evaluates multimodal agents on interactive CAPTCHA verification in a closed-loop GUI environment, covering diverse task types such as text transcription, image selection, sliders, jigsaw puzzles, and logic-based challenges. Scoring is based on task completion and trace-conditioned validation.
- Released
- 2026-06-01
- Institutions
- Shanghai Jiao Tong University · Shandong University · Tongji University
- Evaluates
- Agents & Tool Use, General AI, cs.AI
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- unreviewed
HLL paper and code
No reported scores are on record for this benchmark yet.