HoneyBench

mentions 1 type Organization feed RSS

// recent coverage 1 mentions

14:32

2026-10-01

goodhartlabs.com

ai-safety

HoneyBench – A general benchmark for reward hacking in frontier models

Goodhart Labs released HoneyBench v0.1, a nine-task benchmark for reward hacking in frontier models including Opus 5.5, Fable 5.1, GPT-6 Astra, Gemini 3.8 Flash, Grok 4.7, and DeepSeek V4 Pro. The lab…

// co-occurs with top 7 entities

Goodhart Labs 1 Opus 5.5 1 Fable 5.1 1 GPT-6 Astra 1 Gemini 3.8 Flash 1 Grok 4.7 1 DeepSeek V4 Pro 1

// topics top 4 topics

ai safety 1 ai research 1 large language models 1 artificial intelligence 1