Linear socio-demographic representations emerge in Large Language Models from indirect cues

• URL: http://arxiv.org/abs/2512.10065v1
• Date: Wed, 10 Dec 2025 20:36:36 GMT
• Title: Linear socio-demographic representations emerge in Large Language Models from indirect cues
• Authors: Paul Bouchaud, Pedro Ramaciotti,
• Abstract summary: LLMs encode sociodemographic attributes of human conversational partners inferred from indirect cues such as names and occupations.<n>We show that LLMs develop linear representations of user demographics within activation space, wherein stereotypically associated attributes are encoded along interpretable geometric directions.<n>Our study further highlights that models that pass bias benchmark tests may still harbor and leverage implicit biases, with implications for fairness when applied at scale.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: We investigate how LLMs encode sociodemographic attributes of human conversational partners inferred from indirect cues such as names and occupations. We show that LLMs develop linear representations of user demographics within activation space, wherein stereotypically associated attributes are encoded along interpretable geometric directions. We first probe residual streams across layers of four open transformer-based LLMs (Magistral 24B, Qwen3 14B, GPT-OSS 20B, OLMo2-1B) prompted with explicit demographic disclosure. We show that the same probes predict demographics from implicit cues: names activate census-aligned gender and race representations, while occupations trigger representations correlated with real-world workforce statistics. These linear representations allow us to explain demographic inferences implicitly formed by LLMs during conversation. We demonstrate that these implicit demographic representations actively shape downstream behavior, such as career recommendations. Our study further highlights that models that pass bias benchmark tests may still harbor and leverage implicit biases, with implications for fairness when applied at scale.

Related papers

• Interpretable Debiasing of Vision-Language Models for Social Fairness [55.85977929985967]
  We introduce an interpretable, model-agnostic bias mitigation framework, DeBiasLens, that localizes social attribute neurons in Vision-Language models.<n>We train SAEs on facial image or caption datasets without corresponding social attribute labels to uncover neurons highly responsive to specific demographics.<n>Our research lays the groundwork for future auditing tools, prioritizing social fairness in emerging real-world AI systems.
  arXiv  Detail & Related papers  (2026-02-27T13:37:11Z)
• Overstating Attitudes, Ignoring Networks: LLM Biases in Simulating Misinformation Susceptibility [7.616305266104683]
  Large language models (LLMs) are increasingly used as proxies for human judgment in computational social science.<n>We test whether LLM-simulated survey respondents can reproduce human patterns of misinformation belief and sharing.
  arXiv  Detail & Related papers  (2026-02-04T15:48:05Z)
• Demographic Probing of Large Language Models Lacks Construct Validity [16.29607362682272]
  We study how large language models adapt their behavior to demographic attributes.<n>This approach typically uses a single demographic cue in isolation as a signal for group membership.<n>We find that cues intended to represent the same demographic group induce only partially overlapping changes in model behavior.
  arXiv  Detail & Related papers  (2026-01-26T13:41:35Z)
• A Comprehensive Study of Implicit and Explicit Biases in Large Language Models [1.0555164678638427]
  This study highlights the need to address biases in Large Language Models amid growing generative AI.<n>We studied bias-specific benchmarks such as StereoSet and CrowSPairs to evaluate the existence of various biases in multiple generative models such as BERT and GPT 3.5.<n>Results indicated fine-tuned models struggle with gender biases but excelled at identifying and avoiding racial biases.
  arXiv  Detail & Related papers  (2025-11-18T05:27:17Z)
• Person-Centric Annotations of LAION-400M: Auditing Bias and Its Transfer to Models [81.45743826739054]
  A major barrier has been the lack of demographic annotations in web-scale datasets such as LAION-400M.<n>We create person-centric annotations for the full dataset, including over 276 million bounding boxes, perceived gender and race/ethnicity labels, and automatically generated captions.<n>Using them, we uncover demographic imbalances and harmful associations, such as the disproportionate linking of men and individuals perceived as Black or Middle Eastern with crime-related and negative content.
  arXiv  Detail & Related papers  (2025-10-04T07:51:59Z)
• Reading Between the Prompts: How Stereotypes Shape LLM's Implicit Personalization [13.034294029448338]
  Generative Large Language Models (LLMs) infer user's demographic information from subtle cues in the conversation.<n>Our results highlight the need for greater transparency and control in how LLMs represent user identity.
  arXiv  Detail & Related papers  (2025-05-22T09:48:51Z)
• Beyond Demographics: Fine-tuning Large Language Models to Predict Individuals' Subjective Text Perceptions [33.76973308687867]
  We show that models do improve in sociodemographic prompting when trained.<n>This performance gain is largely due to models learning annotator-specific behaviour rather than sociodemographic patterns.<n>Across all tasks, our results suggest that models learn little meaningful connection between sociodemographics and annotation.
  arXiv  Detail & Related papers  (2025-02-28T09:53:42Z)
• How far can bias go? -- Tracing bias from pretraining data to alignment [54.51310112013655]
  This study examines the correlation between gender-occupation bias in pre-training data and their manifestation in LLMs.<n>Our findings reveal that biases present in pre-training data are amplified in model outputs.
  arXiv  Detail & Related papers  (2024-11-28T16:20:25Z)
• GenderBias-\emph{VL}: Benchmarking Gender Bias in Vision Language Models via Counterfactual Probing [72.0343083866144]
  This paper introduces the GenderBias-emphVL benchmark to evaluate occupation-related gender bias in Large Vision-Language Models. Using our benchmark, we extensively evaluate 15 commonly used open-source LVLMs and state-of-the-art commercial APIs. Our findings reveal widespread gender biases in existing LVLMs.
  arXiv  Detail & Related papers  (2024-06-30T05:55:15Z)
• Leveraging Prototypical Representations for Mitigating Social Bias without Demographic Information [50.29934517930506]
  DAFair is a novel approach to address social bias in language models. We leverage prototypical demographic texts and incorporate a regularization term during the fine-tuning process to mitigate bias.
  arXiv  Detail & Related papers  (2024-03-14T15:58:36Z)
• Sociodemographic Prompting is Not Yet an Effective Approach for Simulating Subjective Judgments with LLMs [13.744746481528711]
  Large Language Models (LLMs) are widely used to simulate human responses across diverse contexts.<n>We evaluate nine popular LLMs on their ability to understand demographic differences in two subjective judgment tasks: politeness and offensiveness.<n>We find that in zero-shot settings, most models' predictions for both tasks align more closely with labels from White participants than those from Asian or Black participants.
  arXiv  Detail & Related papers  (2023-11-16T10:02:24Z)
• On the steerability of large language models toward data-driven personas [98.9138902560793]
  Large language models (LLMs) are known to generate biased responses where the opinions of certain groups and populations are underrepresented. Here, we present a novel approach to achieve controllable generation of specific viewpoints using LLMs.
  arXiv  Detail & Related papers  (2023-11-08T19:01:13Z)
• "Kelly is a Warm Person, Joseph is a Role Model": Gender Biases in LLM-Generated Reference Letters [97.11173801187816]
  Large Language Models (LLMs) have recently emerged as an effective tool to assist individuals in writing various types of content. This paper critically examines gender biases in LLM-generated reference letters.
  arXiv  Detail & Related papers  (2023-10-13T16:12:57Z)
• Bias and Fairness in Large Language Models: A Survey [73.87651986156006]
  We present a comprehensive survey of bias evaluation and mitigation techniques for large language models (LLMs) We first consolidate, formalize, and expand notions of social bias and fairness in natural language processing. We then unify the literature by proposing three intuitive, two for bias evaluation, and one for mitigation.
  arXiv  Detail & Related papers  (2023-09-02T00:32:55Z)

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.