analiza aiqbench
This research introduces CoCoA, a framework that enables large language models to dynamically adjust their entity preferences based on cultural context rather than enforcing uniform neutrality. By utilizing dual-context training and gradient reconciliation, the method significantly reduces cultural bias scores while preserving general model performance across multiple languages. The study establishes a critical new direction for mitigating entity-centric bias in LLMs through context-aware modeling.
CoCoA: Context-Conditional Cultural Alignment for Large Language Models
arXiv:2608.29492v1 Announce Type: new Abstract: Large Language Models (LLMs) often favor Western-associated entities across cultural contexts. Conventional debiasing methods aim for uniform neutrality, but cultural bias mitigation demands context-conditional behavior, preferring culturally appropriate entities when cultural cues are present and remaining neutral when they are absent. We propose CoCoA (Context-Conditional Cultural Alignment), a framework that learns this behavior through dual-context training on the same entity pairs under contexts with and without cultural cues. CoCoA combines a contrastive alignment objective with calibration and drift regularization, optimized through goal-aware gradient reconciliation. We evaluate CoCoA on CAMeL and Camellia, two entity-centric cultural bias benchmarks, across ten language settings and four LLMs. CoCoA reduces the Cultural Bias Score from 43 to 24 on average while maintaining near-neutral preferences at 50.2, with minimal impact on general performance across five standard benchmarks. These findings highlight that effective cultural alignment requires context-conditional modeling rather than uniform debiasing, and establish a new direction for mitigating entity-centric cultural bias in LLMs.
Abstract:Large Language Models (LLMs) often favor Western-associated entities across cultural contexts. Conventional debiasing methods aim for uniform neutrality, but cultural bias mitigation demands context-conditional behavior, preferring culturally appropriate entities when cultural cues are present and remaining neutral when they are absent. We propose CoCoA (Context-Conditional Cultural Alignment), a framework that learns this behavior through dual-context training on the same entity pairs under contexts with and without cultural cues. CoCoA combines a contrastive alignment objective with calibration and drift regularization, optimized through goal-aware gradient reconciliation. We evaluate CoCoA on CAMeL and Camellia, two entity-centric cultural bias benchmarks, across ten language settings and four LLMs. CoCoA reduces the Cultural Bias Score from 43 to 24 on average while maintaining near-neutral preferences at 50.2, with minimal impact on general performance across five standard benchmarks. These findings highlight that effective cultural alignment requires context-conditional modeling rather than uniform debiasing, and establish a new direction for mitigating entity-centric cultural bias in LLMs.
Accepted to Findings of EMNLP 2026. 22 pages, 4 figures, 19 tables
Computation and Language (cs.CL)
arXiv:2608.29492 [cs.CL]
(or arXiv:2608.29492v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2608.29492 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 264 cuvinte
Subiecte conexe
The Continual Constitutional Convention: How language models revise the documents that govern them
AI Alignment Forum · 13 oct. 2026
Scenario-based Probing and Steering Cultural Values in Large Language Models--Extended Version
arXiv · 11 iun. 2026
The J-Space Debate, Agent Swarms, and Pacing Frontier AI - Digital Minds Newsletter #4
LessWrong · 18 sept. 2026
Links #3: 2026/06 Part 1
LessWrong · 15 iun. 2026