Fairness Pruning: Locating Demographic Bias in GLU-MLP Layers via Differential Activations

• URL: http://arxiv.org/abs/2607.28319v1
• Date: Thu, 30 Jul 2026 14:54:33 GMT
• Title: Fairness Pruning: Locating Demographic Bias in GLU-MLP Layers via Differential Activations
• Abstract summary: This work presents Fairness Pruning, a structural intervention method designed for the management and future mitigation of demographic bias in large language models (LLMs)<n>Using minimally contrastive prompt pairs and inference-time activation capture, the method identifies neurons that react differentially when processing demographic attributes in GLU architectures.<n>Results demonstrate that zeroing the identified neurons alters how the model responds to associated demographic variables.
• Score: 0.21410799064827235
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: This work presents Fairness Pruning, a lightweight structural intervention method designed for the management and future mitigation of demographic bias in large language models (LLMs). As a foundational empirical validation of this method, this work focuses on causal bias localization. Using minimally contrastive prompt pairs and inference-time activation capture, the method identifies neurons that react differentially when processing demographic attributes in GLU architectures, evaluating the signal at the down_proj input. Empirical evaluation was conducted on models of up to 3 billion parameters (Llama-3.2 family and Salamandra-2B), combining standardized benchmark evaluation with qualitative text generation experiments. Results demonstrate that zeroing the identified neurons alters how the model responds to associated demographic variables. However, rather than producing flat mitigation, the intervention causes bidirectional bias destabilization: because BiasScore is unsigned, candidate sets mix neurons that push toward and against the stereotype, and the net effect on aggregate bias depends on which sign dominates. The intervention is extremely surgical: zeroing at most 40 neurons in Llama-3.2-1B (less than 0.031% of total MLP width) achieves a mean retention of 99.49% in reasoning and general knowledge capabilities. These findings empirically confirm that demographic bias processing and model capabilities operate on dissociable circuits, establishing the methodological foundations for transitioning from blind zeroing toward directional behavior modulation.

Related papers

• Unraveling Machine Behavior by Multi-Level Bias Analysis and Detection: Methodology and Application to Computer Vision [10.595890561268735]
  This study investigates the presence and propagation of bias within Neural Networks through a comprehensive multi-level analysis.<n>We propose three bias detection approaches: SpaceBias, ActivationBias, and WeightBias.<n>Our proposed techniques provide insight into how biases manifest within the network architecture itself at different levels.
  arXiv  Detail & Related papers  (2026-07-08T10:18:51Z)
• SL-BiLEM: Structured Learnable Behavior-in-the-Loop Epidemic Modeling for Forecasting and Policy Evaluation [42.644852156101685]
  Epidemic forecasting faces a fundamental challenge: human behavior dynamically responds to disease spread.<n>We propose textbfSL-BiLEM (Structured Learnable Behavior-in-the-Loop Epidemic Model), leveraging physical constraints as regularization for robust extrapolation.
  arXiv  Detail & Related papers  (2026-05-26T08:46:04Z)
• Fair outputs, Biased Internals: Causal Potency and Asymmetry of Latent Bias in LLMs for High-Stakes Decisions [0.0]
  We investigate the use of open-weight models for mortgage underwriting using matched applications that differ only in racially-associated names.<n>We show that models show no output-level bias, yet retain and amplify demographic representations across model layers.<n>Through activation steering and novel cross-layer interventions, we demonstrate that this suppressed information is decision-relevant.
  arXiv  Detail & Related papers  (2026-05-12T12:14:58Z)
• Fairness Testing in Retrieval-Augmented Generation: How Small Perturbations Reveal Bias in Small Language Models [0.1876920697241348]
  This study conducts fairness testing through metamorphic testing (MT), introducing controlled demographic perturbations in prompts to assess fairness in sentiment analysis performed by three Small Language Models (SLMs)<n>Results show that minor demographic variations can break up to one third of metamorphic relations (MRs)<n>A detailed analysis of these failures reveals a consistent bias hierarchy, with perturbations involving racial cues being the predominant cause of the violations.
  arXiv  Detail & Related papers  (2025-09-30T17:42:35Z)
• Mitigating Multimodal Hallucinations via Gradient-based Self-Reflection [49.26064449816502]
  We propose a Gradient-based Influence-Aware Constrained Decoding (GACD) method to address text-visual bias and co-occurrence bias.<n>GACD effectively reduces hallucinations and improves the visual grounding of MLLM outputs.
  arXiv  Detail & Related papers  (2025-09-03T08:13:52Z)
• Fair Deepfake Detectors Can Generalize [51.21167546843708]
  We show that controlling for confounders (data distribution and model capacity) enables improved generalization via fairness interventions.<n>Motivated by this insight, we propose Demographic Attribute-insensitive Intervention Detection (DAID), a plug-and-play framework composed of: i) Demographic-aware data rebalancing, which employs inverse-propensity weighting and subgroup-wise feature normalization to neutralize distributional biases; and ii) Demographic-agnostic feature aggregation, which uses a novel alignment loss to suppress sensitive-attribute signals.<n>DAID consistently achieves superior performance in both fairness and generalization compared to several state-of-the-art
  arXiv  Detail & Related papers  (2025-07-03T14:10:02Z)
• NeuronTune: Towards Self-Guided Spurious Bias Mitigation [26.544938760265136]
  Deep neural networks often develop spurious bias, reliance on correlations between non-essential features and classes for predictions.<n>Existing mitigation approaches typically depend on external annotations of spurious correlations.<n>We propose NeuronTune, a post hoc method that directly intervenes in a model's internal decision process.
  arXiv  Detail & Related papers  (2025-05-29T22:33:00Z)
• CausalDiff: Causality-Inspired Disentanglement via Diffusion Model for Adversarial Defense [61.78357530675446]
  Humans are difficult to be cheated by subtle manipulations, since we make judgments only based on essential factors.<n>Inspired by this observation, we attempt to model label generation with essential label-causative factors and incorporate label-non-causative factors to assist data generation.<n>For an adversarial example, we aim to discriminate perturbations as non-causative factors and make predictions only based on the label-causative factors.
  arXiv  Detail & Related papers  (2024-10-30T15:06:44Z)
• Mitigating Biases for Instruction-following Language Models via Bias Neurons Elimination [54.865941973768905]
  We propose a novel and practical bias mitigation method, CRISPR, to eliminate bias neurons of language models in instruction-following settings. CRISPR automatically determines biased outputs and categorizes neurons that affect the biased outputs as bias neurons using an explainability method. Experimental results demonstrate the effectiveness of our method in mitigating biases under zero-shot instruction-following settings without losing the model's task performance and existing knowledge.
  arXiv  Detail & Related papers  (2023-11-16T07:16:55Z)
• Balancing Biases and Preserving Privacy on Balanced Faces in the Wild [50.915684171879036]
  There are demographic biases present in current facial recognition (FR) models. We introduce our Balanced Faces in the Wild dataset to measure these biases across different ethnic and gender subgroups. We find that relying on a single score threshold to differentiate between genuine and imposters sample pairs leads to suboptimal results. We propose a novel domain adaptation learning scheme that uses facial features extracted from state-of-the-art neural networks.
  arXiv  Detail & Related papers  (2021-03-16T15:05:49Z)
• LOGAN: Local Group Bias Detection by Clustering [86.38331353310114]
  We argue that evaluating bias at the corpus level is not enough for understanding how biases are embedded in a model. We propose LOGAN, a new bias detection technique based on clustering. Experiments on toxicity classification and object classification tasks show that LOGAN identifies bias in a local region.
  arXiv  Detail & Related papers  (2020-10-06T16:42:51Z)
• An Investigation of Why Overparameterization Exacerbates Spurious Correlations [98.3066727301239]
  We identify two key properties of the training data that drive this behavior. We show how the inductive bias of models towards "memorizing" fewer examples can cause over parameterization to hurt.
  arXiv  Detail & Related papers  (2020-05-09T01:59:13Z)

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.