tyrtleli/thinking-benchmark-gpt5 download history

tyrtleli/thinking-benchmark-gpt5 is a question answering dataset on the Hugging Face Hub. In the last 30 days it was downloaded 14 times (1 in the last 7 days), and 105 times in total. It ranks #470,174 among datasets by monthly downloads.

Thinking Benchmark — GPT-5 k=32 subset Five problems selected for the k=32 deep pass of the Cost of Overthinking study, targeting GPT-5's capability edge (25–75% success rate on the k=8 calibration pass). This is a subset of tyrtleli/thinking-benchmark. See the parent dataset for full

Open tyrtleli/thinking-benchmark-gpt5 on Hugging Face

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