josefchen/co-failure-67-models download history

josefchen/co-failure-67-models is an other dataset on the Hugging Face Hub. In the last 30 days it was downloaded 644 times (36 in the last 7 days), and 1,103 times in total. It ranks #27,053 among datasets by monthly downloads.

Co-Failure Matrices: 67 LLMs Per-model, per-question outcome data, raw model responses and audit files behind Combining LLMs Rarely Beats the Single Best Model: A Provable Co-Failure Ceiling Across 67 Frontier Models (Josef Chen, KAIKAKU; arXiv:2606.27288). Paper: https://arxiv.org/a

Spaces using co-failure-67-models

• Combining LLMs Rarely Beats the Single Best Model 1 likes

Open josefchen/co-failure-67-models on Hugging Face