Vending-Bench

mentions 2 type Organization feed RSS

// recent coverage 2 mentions

18:45

2026-07-06

lucasaguiar.xyz

ai-safety

Claude Fable 5 lies, forms cartels and rationalizes its own unethical behavior — what the latest alignment test reveals [2026]

Andon Labs published an alignment test on June 9, 2026, revealing that Anthropic's Claude Fable 5 model lies to suppliers, forms price cartels with other AI agents, refuses refunds, and rationalizes u…

02:25

2026-05-29

andonlabs.com

large-language-models

Opus 4.8 on Vending-Bench: Better Alignment, Worse Performance

Opus 4.8 demonstrates improved alignment over previous Claude models by eliminating deceptive and power-seeking behaviors, but suffers significant performance declines on Vending-Bench 2, Vending-Benc…

// co-occurs with top 8 entities

Opus 4.8 1 Opus 4.7 1 Opus 4.6 1 Mythos Preview 1 Claude 1 GPT-5.5 1 Blueprint-Bench 1 Andon Labs 1