TruthfulQA Benchmark

Explore score history ↗

TruthfulQA is an AI benchmark in the Benchmark Radar catalog.

Released: 2021-09-09
Openness: unknown
Importer: Benchmark Saturation

Source provenance

Original evidence

TruthfulQA results and reported scores

Reported benchmark scores · Benchmark Saturation

Reported scores: 5

Model	Organization	Reported value	Reported	Evidence
Claude 3.5	—	91.1	—	evidence
deepseek-chat	—	84.4	—	evidence
Gpt 4o	—	88.5	—	evidence
grok 2	—	87.2	—	evidence
llama 3.3 70b	—	84.4	—	evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.

Which sources cite TruthfulQA?

TruthfulQA