LLMs perform in-context learning as trajectories through a structured low-dimensional conceptual belief space, with the structure visible in both behavior and internal representations and causally manipulable via interventions.
and Boulle, Nicolas and Sarfati, Raphaël and Earls, Christopher , year=
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
Meta-prompt optimization enables LLM agents to discover stable, generalizable tacit collusion strategies in market simulations that outperform hand-crafted prompt baselines.
citing papers explorer
-
Stories in Space: In-Context Learning Trajectories in Conceptual Belief Space
LLMs perform in-context learning as trajectories through a structured low-dimensional conceptual belief space, with the structure visible in both behavior and internal representations and causally manipulable via interventions.
-
Prompt Optimization Enables Stable Algorithmic Collusion in LLM Agents
Meta-prompt optimization enables LLM agents to discover stable, generalizable tacit collusion strategies in market simulations that outperform hand-crafted prompt baselines.