Analyzing and Correcting Benevolence Bias in Large Language Models

• URL: http://arxiv.org/abs/2608.24912v1
• Date: Sun, 26 Jul 2026 08:37:58 GMT
• Title: Analyzing and Correcting Benevolence Bias in Large Language Models
• Abstract summary: Large language models (LLMs) are increasingly used as stand-ins for human respondents.<n>Here we identify and measure benevolence bias, a small but consistent tendency for aligned LLMs to lean toward the kinder, safer, more socially approved answer on value-laden survey questions.<n>We find that the bias is a stable model property, not a quirk of any one system.
• Score: 25.94213078450494
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models (LLMs) are increasingly used as stand-ins for human respondents, from opinion polls and simulated survey participants to agent-based social simulations. These uses rest on one assumption: that conditioning a model on who a person is yields answers resembling those of real people from that group. Here we identify and measure benevolence bias, a small but consistent tendency for aligned LLMs to lean toward the kinder, safer, more socially approved answer on value-laden survey questions. Across 18 widely used models, four social-science datasets (ANES, GSS, WVS, and a cross-cultural prospect-theory replication) and six psychological categories, we find that the bias is a stable model property, not a quirk of any one system: it points the same way across models, grows with model size, and traces to the post-training stage. Prompt language and framing change its size but never its direction, and a "malicious persona" stress test shows a one-sided limit: aligned models struggle to play people who are less kind, less prosocial or more harm-tolerant than average. The issue is thus not only a shifted average, but a narrowed range of people the model can imitate. The bias sits in the middle of the answer distribution rather than its tails, and survives changes in sampling temperature and simple prompted reflection. The encouraging news is that it is easy to diagnose and straightforward to fix: a light-touch contrastive calibration, which needs no retraining and works on black-box APIs, brings all six categories back to the human baseline. Our results give researchers a clear map of where aligned LLMs can already be trusted as human stand-ins, where they need care, and a ready-to-use method for closing the gap.

Related papers

• Mind the Gaps: Mixture-of-Minds for Human Simulation [1.1844977816228046]
  Anacreon is an audience simulation model that targets the individual level within a narrow, well-specified domain.<n>It harvests demographics, psychological traits, and survey responses from public text, and augments each record with a chain-of-emotion.<n>It reduces prompt brittleness by shuffling response options and reduces positive bias by balancing the training distribution.
  arXiv  Detail & Related papers  (2026-08-06T14:48:09Z)
• The Chameleon's Limit: Investigating Persona Collapse and Homogenization in Large Language Models [12.654067061944074]
  Applications based on large language models (LLMs) require population diversity among agents.<n>We identify a pervasive failure mode we term emphPersona Collapse: agents each assigned a distinct profile nonetheless converge into a narrow behavioral mode, producing a homogeneous simulated population.
  arXiv  Detail & Related papers  (2026-04-27T17:01:48Z)
• Surface Fairness, Deep Bias: A Comparative Study of Bias in Language Models [45.41676783204022]
  We investigate various proxy measures of bias in large language models (LLMs)<n>We find that evaluating models with pre-prompted personae on a multi-subject benchmark (MMLU) leads to negligible and mostly random differences in scores.<n>With the recent trend for LLM assistant memory and personalization, these problems open up from a different angle.
  arXiv  Detail & Related papers  (2025-06-12T08:47:40Z)
• The Devil is in the Neurons: Interpreting and Mitigating Social Biases in Pre-trained Language Models [78.69526166193236]
  Pre-trained Language models (PLMs) have been acknowledged to contain harmful information, such as social biases. We propose sc Social Bias Neurons to accurately pinpoint units (i.e., neurons) in a language model that can be attributed to undesirable behavior, such as social bias. As measured by prior metrics from StereoSet, our model achieves a higher degree of fairness while maintaining language modeling ability with low cost.
  arXiv  Detail & Related papers  (2024-06-14T15:41:06Z)
• Large Language Models Show Human-like Social Desirability Biases in Survey Responses [12.767606361552684]
  We show that Large Language Models (LLMs) skew their scores towards the desirable ends of trait dimensions when personality evaluation is inferred. This bias exists in all tested models, including GPT-4/3.5, Claude 3, Llama 3, and PaLM-2. reverse-coding all the questions decreases bias levels but does not eliminate them, suggesting that this effect cannot be attributed to acquiescence bias.
  arXiv  Detail & Related papers  (2024-05-09T19:02:53Z)
• Self-Debiasing Large Language Models: Zero-Shot Recognition and Reduction of Stereotypes [73.12947922129261]
  We leverage the zero-shot capabilities of large language models to reduce stereotyping. We show that self-debiasing can significantly reduce the degree of stereotyping across nine different social groups. We hope this work opens inquiry into other zero-shot techniques for bias mitigation.
  arXiv  Detail & Related papers  (2024-02-03T01:40:11Z)
• ROBBIE: Robust Bias Evaluation of Large Generative Language Models [27.864027322486375]
  Different prompt-based datasets can be used to measure social bias across multiple text domains and demographic axes. We compare 6 different prompt-based bias and toxicity metrics across 12 demographic axes and 5 families of generative LLMs. We conduct a comprehensive study of how well 3 bias/toxicity mitigation techniques perform across our suite of measurements.
  arXiv  Detail & Related papers  (2023-11-29T23:03:04Z)
• Do LLMs exhibit human-like response biases? A case study in survey design [66.1850490474361]
  We investigate the extent to which large language models (LLMs) reflect human response biases, if at all. We design a dataset and framework to evaluate whether LLMs exhibit human-like response biases in survey questionnaires. Our comprehensive evaluation of nine models shows that popular open and commercial LLMs generally fail to reflect human-like behavior.
  arXiv  Detail & Related papers  (2023-11-07T15:40:43Z)
• Non-Invasive Fairness in Learning through the Lens of Data Drift [88.37640805363317]
  We show how to improve the fairness of Machine Learning models without altering the data or the learning algorithm. We use a simple but key insight: the divergence of trends between different populations, and, consecutively, between a learned model and minority populations, is analogous to data drift. We explore two strategies (model-splitting and reweighing) to resolve this drift, aiming to improve the overall conformance of models to the underlying data.
  arXiv  Detail & Related papers  (2023-03-30T17:30:42Z)
• Spuriosity Rankings: Sorting Data to Measure and Mitigate Biases [62.54519787811138]
  We present a simple but effective method to measure and mitigate model biases caused by reliance on spurious cues. We rank images within their classes based on spuriosity, proxied via deep neural features of an interpretable network. Our results suggest that model bias due to spurious feature reliance is influenced far more by what the model is trained on than how it is trained.
  arXiv  Detail & Related papers  (2022-12-05T23:15:43Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.