FRAMES

text
+
+
+
+
About

FRAMES (Factuality, Retrieval, And reasoning MEasurement Set) is a comprehensive evaluation dataset designed to test Retrieval-Augmented Generation systems' capabilities. Created by Google, this benchmark measures AI models' factuality, retrieval accuracy, and reasoning abilities in RAG contexts. FRAMES evaluates how well models can retrieve relevant information and generate factually accurate responses based on retrieved content.

+
+
+
+
Evaluation Stats
Total Models1
Organizations1
Verified Results0
Self-Reported1
+
+
+
+
Benchmark Details
Max Score1
Language
en
+
+
+
+
Performance Overview
Score distribution and top performers

Score Distribution

1 models
Top Score
73.3%
Average Score
73.3%
High Performers (80%+)
0

Top Organizations

#1DeepSeek
1 model
73.3%
+
+
+
+
Leaderboard
1 models ranked by performance on FRAMES
LicenseLinks
Dec 25, 2024
MIT + Model License (Commercial use allowed)
73.3%
+
+
+
+
Resources