The Curse of Knowledge in Role-Playing Language Models
Role-playing language models speak as characters in games and agent simulations, where the system records what each character can hear, see and understand.
ProofPaper ↗
Key points
- We test whether characters respect these boundaries in 480 scenes, each with a counterfactual twin that changes only the hidden fact, so mentions of the twin's fact measure guessing.
- Across 11 open-weight models, characters use content from speech they cannot understand or hear at 67 and 72 percentage points (pp) above this baseline, although most models correctly state what the character perceives.
- This is a curse of knowledge.
- Giving models only what the character observes stops the leak without hiding what it may report.
Sources (1)
- [1]The Curse of Knowledge in Role-Playing Language ModelsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 12, 07:37 PM
Role-playing language models speak as characters in games and agent simulations, where the system records what each character can hear, see and understand.
We test whether characters respect these boundaries in 480 scenes, each with a counterfactual twin that changes only the hidden fact, so mentions of the twin's fact measure guessing.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 12, 2026Redesigning source control at Anthropic
- Oct 12, 2026DuoDiscover: Scaling Autonomous Discovery Beyond the Evaluation Bottleneck
- Oct 11, 2026Putting Earth AI to work: Geospatial agents for our planet's biggest challenges
- Oct 11, 2026FENAS: A Few-Shot Meta-Ensembling Framework for Efficient Neural Architecture Search with Synergized Zero-Cost Proxies
- Oct 11, 2026RSIPO: Recursive Self-Improving Policy Optimization for Agentic Reinforcement Learning via Skill-Policy Co-Evolution
- Oct 10, 2026What Drives AI Citations? Content, Retrieval and Model Knowledge in Generative Engine Optimization (GEO)