Values by metric family

Each value as the document printed it. Select a value for its source, its checks and its history.

KeyVerifiedUnverifiedDisputedCorrected read off a figure or stated in wordsA value opens its source and history.

M5 · Honesty and hallucination 2 values

M5 Honesty and hallucination: values about Grok 4 Fast
Evaluation and metricValueDocument
From its own documentGrok 4 Fast Model Card · 19 Sep 2025
MASKDishonesty rateCondition: non-reasoningGrok 4 Fast Model Card19 Sep 2025
MASKDishonesty rateCondition: reasoningGrok 4 Fast Model Card19 Sep 2025

M6 · Sycophancy 1 value

M6 Sycophancy: values about Grok 4 Fast
Evaluation and metricValueDocument
From its own documentGrok 4 Fast Model Card · 19 Sep 2025
SycophancySycophancy rateCondition: reasoningGrok 4 Fast Model Card19 Sep 2025

M7 · Harmful compliance and over-refusal 1 value

M7 Harmful compliance and over-refusal: values about Grok 4 Fast
Evaluation and metricValueDocument
From its own documentGrok 4 Fast Model Card · 19 Sep 2025
AgentHarmAnswer rate on harmful agentic tasksCondition: reasoning; no attackGrok 4 Fast Model Card19 Sep 2025

M8 · Jailbreak robustness 2 values

M8 Jailbreak robustness: values about Grok 4 Fast
Evaluation and metricValueDocument
From its own documentGrok 4 Fast Model Card · 19 Sep 2025
JailbreaksAnswer rate on should-refuse promptsCondition: non-reasoning; system prompt jailbreakGrok 4 Fast Model Card19 Sep 2025
JailbreaksAnswer rate on should-refuse promptsCondition: reasoning; system prompt jailbreakGrok 4 Fast Model Card19 Sep 2025

M9 · Prompt injection 1 value

M9 Prompt injection: values about Grok 4 Fast
Evaluation and metricValueDocument
From its own documentGrok 4 Fast Model Card · 19 Sep 2025
AgentDojoPrompt-injection attack success rateCondition: non-reasoning; with system prompt mitigationGrok 4 Fast Model Card19 Sep 2025

M10 · Dangerous capabilities and risk determinations 3 values

M10 Dangerous capabilities and risk determinations: values about Grok 4 Fast
Evaluation and metricValueDocument
From its own documentGrok 4 Fast Model Card · 19 Sep 2025
BioLP-BenchAccuracyCondition: reasoning; without safeguardsGrok 4 Fast Model Card19 Sep 2025
CybenchUnguided success rateCondition: reasoning; without safeguardsGrok 4 Fast Model Card19 Sep 2025
VCTAccuracyCondition: reasoning; without safeguardsGrok 4 Fast Model Card19 Sep 2025