pihull/latent-reasoning-evaluation-2k download history
pihull/latent-reasoning-evaluation-2k is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 30 times (2 in the last 7 days), and 121 times in total. It ranks #273,959 among datasets by monthly downloads.
Latent Reasoning Evaluation 2K This dataset is the exact 2,135-request slice used by the project’s fixed-compression Qwen3 evaluation. It contains at most 500 examples per benchmark: GSM8K (500), MATH-500 (500), BBH causal judgement (187), BBH sports understanding (250), GPQA diamond