← Back to all news
arXiv AI By Axel Abels, Elias Fern\'andez Domingos, Tom Lenaerts

Jointly Modeling Preferences and Imperfect Decision-Making in Large Language Models

Read the original on arXiv AI →

The paper introduces the Preference-Behavior Model (PBM), which jointly infers Large Language Model (LLM) preferences and imperfect decision-making mechanisms from observed choices. By applying PBM to three social dilemma games—Public Goods, Collective-Risk, and Common-Pool Resource—the authors show that preferences derived from one dilemma can predict behavior in others, whereas traditional behavioral economic elicitation instruments fail to do so consistently. The study also evaluates how well inferred preferences align with persona instructions, revealing both broad agreement and specific divergences across dimensions, LLMs, and games.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Related stories

arXiv AI
Aug 20

How AI Prompts Can Teach Us About the Structure of Human Behavior

The paper presents an AI-based method that uses a large language model to emulate human decision-making by assigning it a "type vector" describing traits such as Altruism and Risk Aversion. By varying these dimensions and values, the authors fit the model to over 119,000 decisions from 78,657 participants in 10 classic economic games, finding that three dimensions—Risk Aversion, Strategic Sophistication, and Trust—sufficiently capture human behavior. The resulting type clusters, fewer than a dozen, predict behavior in new games, suggesting a low-dimensional, portable representation of human behavior across diverse settings.

By Matthew O. Jackson, Benjamin S. Manning, Yutong Xie, Walter Yuan, Qiaozhu Mei
llmssafety
More like this →
arXiv Computation and Language
Oct 7

Large Language Model Orchestration under Heterogeneous Preferences via Explicit Persona Inference

The paper introduces HARP, a framework for orchestrating heterogeneous agents with hidden preferences by maintaining numeric posterior beliefs updated via Bayes’ rule, rather than embedding beliefs in prompts. HARP achieves ∼O(√K) Bayesian regret and, with the HARP+ variant, adds a bonus for informative actions to keep inference active even when optimal actions are uninformative. Experiments across three problem settings show HARP+ outperforms other non‑oracle methods in scenarios where explicit joint inference is infeasible.

By Shuqing Shi, Ziyan Wang, Milind Tambe, Yali Du
llmsagents
More like this →