Skip to content

Dataset · google

Google benchmark for chain-of-thought verification

google/reveal

English dataset for evaluating verifier models that check step-by-step LLM reasoning.

What's inside

This is a benchmark from Google for evaluating the quality of Chain-of-Thought verification in language models. Inside are 1 to 10 thousand English examples presented in text and tabular formats. The dataset was created for text classification and question-answering tasks.

What you can do with it

  • Evaluating verifier models that search for errors in step-by-step LLM reasoning
  • Testing answer reliability control systems in question-answering tasks
Tasks
Text classification, Question answering
Languages
English
Size
1K–10K rows
Downloads this month
376
Likes
40
Access
on request on Hugging Face

How to load it

from datasets import load_dataset

ds = load_dataset("google/reveal")