GPT-3 175B (0-shot)

Reported on 6 benchmarks across 3 tasks · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing6 results

Reading ComprehensiononRACE
Accuracy (Middle)· 2020-05-28
58.4
best: 93.1 (Megatron-BERT (ensemble))
Language Models are Few-Shot Learners arXiv:2005.14165
Question AnsweringonCOPA
Accuracy· 2020-05-28
91
best: 100 (PaLM 540B (finetuned) )
Language Models are Few-Shot Learners arXiv:2005.14165
Question AnsweringonPIQA
Accuracy· 2020-05-28
81
best: 90.1 (Unicorn 11B (fine-tuned))
Language Models are Few-Shot Learners arXiv:2005.14165
Common Sense ReasoningonWinoGrande
Accuracy· 2020-05-28
70.2
best: 96.1 (ST-MoE-32B 269B (fine-tuned))
Language Models are Few-Shot Learners arXiv:2005.14165
Common Sense ReasoningonARC (Challenge)
Accuracy· 2020-05-28
51.4
best: 96.4 (GPT-4 (few-shot, k=25))
Language Models are Few-Shot Learners arXiv:2005.14165
Common Sense ReasoningonARC (Easy)
Accuracy· 2020-05-28
68.8
best: 95.2 (ST-MoE-32B 269B (fine-tuned))
Language Models are Few-Shot Learners arXiv:2005.14165