리플레이: https://hehee9.github.io/maze-bench/public/index.html리더보드: https://hehee9.github.io/maze-bench/public/leaderboard.htmlHugging Face Space: https://huggingface.co/spaces/Hehee-dev/maze-benchGitHub: https://github.com/hehee9/maze-bench (제작: hehee9, Apache 2.0)바로가기 Maze Bench Replay — Compare Model Maze-Solving PathsReplay and compare how LLMs navigate visual mazes, including optimal paths, m..
'Benchmark' 태그의 글 목록