Newest papers

Effects of Model Scale and Quantization on Social Bias and Abstention in Question Answering

Kuniko Paxton, Ifeoluwa Wuraola, Dhavalkumar Thakker · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.29657 · cited by 0 works (OpenAlex) · open-access PDF

French

Le déploiement de modèles de langue petits et quantifiés dans des réglages limités par les ressources soulève une question d’équité importante : quand un modèle comprimé produit moins de réponses biaisées, cela reflète-t-il un comportement plus équitable ou une plus grande abstinence ? Nous étudions ce problème dans Question Answering (QA) en utilisant le benchmark BBQ sur plusieurs tailles de modèle et les réglages de quantification de Qwen3, Gemma-3 et SmolLM2. Nous examinons également si les tendances observées restent stables dans les formulations de prompt alternatives et si l’option « inconnu » est supprimée dans un cadre de choix forcé. Nos résultats montrent que les modèles plus grands ont atteint une précision QA plus élevée

Machine translation (Meta M2M100); the English original follows.

Abstract

The deployment of small and quantized language models in resource-constrained settings raises an important fairness question: when a compressed model produces fewer biased responses, does this reflect fairer behavior or greater abstention? We study this problem in Question Answering (QA) using the BBQ benchmark across multiple model sizes and quantization settings from Qwen3, Gemma-3, and SmolLM2. We also examine whether the observed trends remain stable across alternative prompt formulations and when the ``Unknown'' option is removed in a forced-choice setting. Our results show that larger models achieved higher QA accuracy, but bias showed no simple scaling trend and quantization effects varied across configurations. Notably, 20 of the 26 configurations with fewer biased responses also show increased over-abstention, indicating that lower observed bias can partly reflect greater abstention. These findings show that lower observed bias can partly reflect greater abstention, and that fairness evaluation should consider biased responses, abstention, prompt formulation, and answer-option design jointly.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Quantization (signal processing), Scaling, Question answering, Econometrics, Computer science, Psychology

Is this paper real? This page was built from OpenAlex and arXiv records, not generated. Software can verify any reference in one call: GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.29657, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txt

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange