B2J/hallumaze-benchmark download history

B2J/hallumaze-benchmark is a text generation dataset on the Hugging Face Hub. In the last 30 days it was downloaded 83 times (53 in the last 7 days), and 342 times in total. It ranks #129,953 among datasets by monthly downloads.

HalluMaze Benchmark Dataset All 13 tested LLMs score significantly below a random walk on metacognitive recovery (p<0.001). Newer ≠ better: Claude-Sonnet-4.5 (MEI=0.783) surpasses 4.6 (MEI=0.545). SR and MEI are orthogonal: 4.6 has highest SR (60%) but ranks 8th. Dataset

Open B2J/hallumaze-benchmark on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.