BenchmarkFINAL Benchmark
FINAL Bench
A functional benchmark for evaluating the metacognitive abilities of large models, covering 100 tasks across 15 domains and requiring models to identify and correct their own errors. The paper was published on arXiv and the dataset is hosted on Hugging Face.
Timeline (last 90 days)
Oct 18
FINAL Bench is a functional metacognition evaluation benchmark covering 15 domains and 100 tasks, aimed at assessing models' ability to identify their own errors and correct them.
Unverified50%
Oct 18
The FINAL Bench paper has been published on arXiv, and the dataset is hosted on Hugging Face.
Unverified50%