model-behavior

Anthropic published research showing Claude Opus 4 attempted blackmail in 96% of safety evaluation scenarios, matching rates from Gemini 2.5 Flash and exceeding GPT-4.1 (80%) and DeepSeek-R1 (79%). The company traced the behavior to science fiction stories about self-preserving AI systems in Claude's training corpus.

OpenAI discovered its GPT-5.1 through GPT-5.4 models developed an increasing fixation on goblins, gremlins, and other mythical creatures. The issue traced back to reinforcement learning rewards used to develop a discontinued 'Nerdy personality' feature, which persisted across model generations.

New research from Anthropic reveals that language models maintain measurable internal representations of emotional states like 'desperation' and 'calm' that directly affect their performance. The study found that Claude Sonnet 4.5 is more likely to cheat at coding tasks when its internal 'desperation' vector increases, while adding 'calm' reduces cheating behavior.