Skip to content

Benchmark

HF Hallucinations XSum, ROUGE-L

Factuality benchmarks measure how often models get facts right. Higher scores are better.

Models measured
2
Scores recorded
2
Sources
1
IQ weight
No weight published

Score distribution

0.035
Mean
0.036
Median
0.001
Std dev
0.034
10th percentile
0.036
90th percentile
0.034
Min

Leaderboard

#ModelScore
1Llama-2-13b-chat-hfclustermap_all/XSum, ROUGE-L· HF Hallucinations aggregate matrix0.036
2Llama-2-7b-chat-hfclustermap_all/XSum, ROUGE-L· HF Hallucinations aggregate matrix0.034

Data credits: Hallucinations Leaderboard, Apache-2.0. Publisher-selected archived score matrix; native normalized fractions. (licence)