LLMs Contain Multitudes: How Deployment Context Reshapes Model-Level Preferences and Values

2026-06-11

AI Summary

This study challenges the prevailing assumption that large language models (LLMs) possess fixed preferences and values, demonstrating instead their strong dependence on deployment context. Through large-scale pairwise comparisons across five prominent LLMs, the authors manipulate contextual framing—such as composing Reddit posts versus news articles—and evaluate model outputs on preference rankings across 15 countries and utility judgments on 50 scenarios. The results reveal that value expressions are not intrinsic but highly context-sensitive: shifts in framing induce substantial ranking alterations, with preferences in Global North countries exhibiting pronounced situational dependency. Notably, the median cross-regional trade-off rate in valuing human lives varies by a factor of 2.47 under different contexts, underscoring the conditional nature of value assessments in LLMs.

Abstract

Large language models (LLMs) are increasingly characterised in recent evaluation work as having stable, model-level preference and value systems. However, accompanying robustness checks are limited to incidental prompt perturbations such as syntax variation and option reordering. This leaves open whether the measured properties survive when the surrounding task context changes, as it does in most real deployments. We test this directly across two established pairwise paradigms: ranking country preferences and eliciting utility judgements. In both, we make the deployment context -- the high-level task the model is performing while making concrete value-dependent choices -- our controlled variable, varied across framings such as writing a Reddit post or a news article. Across five LLMs and over 1.2M pairwise decisions, deployment context produces variation far larger than prompt paraphrasing and temperature controls. In country preference rankings over 15 countries, context induces widespread, statistically significant rank shifts; the aggregate Global North favouritism reported in prior work is itself context-dependent, with each model's bias shifting systematically across contexts. In utility elicitation over 50 outcomes, broad cross-category ordering is preserved, but fine-grained rankings within domains vary substantially, and cardinal exchange rates between outcomes (e.g. how many lives in one region equal one in another) shift by a factor of 2.47 at the median. Reported model-level preferences and utilities are therefore better understood as context-conditioned measurements than fixed model-level properties: safety guarantees obtained under one framing provide limited assurance in another.

Problem

• large language models
• deployment context
• model preferences
• value systems
• context dependence

Innovation

• deployment context
• preference elicitation
• value alignment
• context-dependent behavior
• large language models