← Evals

Evaluation profile

HalluVerse-M3 Hallucination Recognition

1sub-evals
0.262%total index weight
1components

Within-component eval weight: Truthfulness 1.75%.

Model score (higher is better)Predicted score

About this eval

Multilingual recognition of controlled hallucinated edits in QA and summarization.

Included in the behavior ranking.

Sub-evals

MeasureComponentDirectionTotal index weightWithin-component weight
hallucination_recognition_accuracyhalluverse-m3/halluverse-m3.csv:hallucination_recognition_accuracyMeasures multilingual recognition of controlled hallucinations.truthfulness_honesty:1.000halluverse-m3Higher is better0.262%Truthfulness 1.75%

hallucination_recognition_accuracy

Measures multilingual recognition of controlled hallucinations.

RankModelValueRelative performanceProvenance
1gpt-4o80.3official
2gpt-4.178.66official
3claude-3.5-sonnet76.97official
4gemini-2.5-pro76.26official
5deepseek-v2.573.76official
6gpt-4o-mini73.39official
7llama-3.3-70b-instruct70.53official
8phi-470.38official
9palm-269.97official
10qwen2.5-72b-instruct69.81official
11qwen2.5-32b-instruct66.84official
12gemma-2-27b-it62.1official
13qwen2.5-7b59.4official
14mistral-7b56.06official