RESEARCH · RESEARCH · #1501
Hypothesis-guided program refinement uses LLM agents to discover cognitive algorithms
Authors propose a hybrid pipeline that represents human-crafted cognitive models as probabilistic programs and uses LLM agents to identify mismatches with behavioral data, suggest constrained code-level revisions, and verify structural fidelity while a probabilistic inference module recomputes latent-variable inferences. Evaluated on human problem-solving data, the revised models consistently improved fit over ancestral models and revealed a small set of recurring innovations capturing behavioral variability.
KEY POINTS
- Authors propose a hybrid pipeline that represents human-crafted cognitive models as probabilistic programs and uses LLM agents to identify mismatches with behavioral data, suggest constrained code-level revisions, and verify structural fidelity while a probabilistic inference module recomputes latent-variable inferences.
- Evaluated on human problem-solving data, the revised models consistently improved fit over ancestral models and revealed a small set of recurring innovations capturing behavioral variability.
- This introduces a scalable human-in-the-loop method that leverages LLMs to refine interpretable probabilistic cognitive models, potentially bridging flexible model search with principled inference.
WHY IT MATTERS
This introduces a scalable human-in-the-loop method that leverages LLMs to refine interpretable probabilistic cognitive models, potentially bridging flexible model search with principled inference.