Aligning Attention with Human Rationales for Self-Explaining Hate Speech Detection

• URL: http://arxiv.org/abs/2511.07065v1
• Date: Mon, 10 Nov 2025 12:57:56 GMT
• Title: Aligning Attention with Human Rationales for Self-Explaining Hate Speech Detection
• Authors: Brage Eilertsen, Røskva Bjørgfinsdóttir, Francielle Vargas, Ali Ramezani-Kebrya,
• Abstract summary: Supervised Rational Attention (SRA) is a framework that explicitly aligns model attention with human rationales.<n>SRA improves both interpretability and fairness in hate speech classification.
• Score: 2.5432391525687748
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The opaque nature of deep learning models presents significant challenges for the ethical deployment of hate speech detection systems. To address this limitation, we introduce Supervised Rational Attention (SRA), a framework that explicitly aligns model attention with human rationales, improving both interpretability and fairness in hate speech classification. SRA integrates a supervised attention mechanism into transformer-based classifiers, optimizing a joint objective that combines standard classification loss with an alignment loss term that minimizes the discrepancy between attention weights and human-annotated rationales. We evaluated SRA on hate speech benchmarks in English (HateXplain) and Portuguese (HateBRXplain) with rationale annotations. Empirically, SRA achieves 2.4x better explainability compared to current baselines, and produces token-level explanations that are more faithful and human-aligned. In terms of fairness, SRA achieves competitive fairness across all measures, with second-best performance in detecting toxic posts targeting identity groups, while maintaining comparable results on other metrics. These findings demonstrate that incorporating human rationales into attention mechanisms can enhance interpretability and faithfulness without compromising fairness.

Related papers

• Stable Language Guidance for Vision-Language-Action Models [62.80963701282789]
  Residual Semantic Steering is a probabilistic framework that disentangles physical affordance from semantic execution.<n> RSS achieves state-of-the-art robustness, maintaining performance even under adversarial linguistic perturbations.
  arXiv  Detail & Related papers  (2026-01-07T16:16:10Z)
• Self-Explaining Hate Speech Detection with Moral Rationales [11.165386773222934]
  We propose Supervised Moral Rationale Attention (SMRA), the first self-explaining hate speech detection framework to incorporate moral rationales as direct supervision for attention alignment.<n>Based on Moral Foundations Theory, SMRA aligns token-level attention with expert-annotated moral rationales, guiding models to attend to morally salient spans rather than spurious lexical patterns.
  arXiv  Detail & Related papers  (2026-01-07T00:17:16Z)
• Confident, Calibrated, or Complicit: Probing the Trade-offs between Safety Alignment and Ideological Bias in Language Models in Detecting Hate Speech [0.916708284510944]
  We investigate the efficacy of Large Language Models (LLMs) in detecting implicit and explicit hate speech.<n>We find that censored models significantly outperform their uncensored counterparts in both accuracy and robustness.
  arXiv  Detail & Related papers  (2025-08-31T03:00:55Z)
• 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)
• Aligning ASR Evaluation with Human and LLM Judgments: Intelligibility Metrics Using Phonetic, Semantic, and NLI Approaches [28.79400870481616]
  We identify two key challenges: (1) Existing metrics do not adequately reflect intelligibility, and (2) while LLMs can refine ASR output, their effectiveness in correcting ASR transcripts is underexplored.<n>We propose a novel metric integrating Natural Language Inference (NLI) scores, semantic similarity, and phonetic similarity.<n>Our ASR evaluation metric achieves a 0.890 correlation with human judgments on Speech Accessibility Project data, surpassing traditional methods and emphasizing the need to prioritize intelligibility over error-based measures.
  arXiv  Detail & Related papers  (2025-06-19T18:21:19Z)
• Retrieval is Not Enough: Enhancing RAG Reasoning through Test-Time Critique and Optimization [58.390885294401066]
  Retrieval-augmented generation (RAG) has become a widely adopted paradigm for enabling knowledge-grounded large language models (LLMs)<n>RAG pipelines often fail to ensure that model reasoning remains consistent with the evidence retrieved, leading to factual inconsistencies or unsupported conclusions.<n>We propose AlignRAG, a novel iterative framework grounded in Critique-Driven Alignment (CDA)<n>We introduce AlignRAG-auto, an autonomous variant that dynamically terminates refinement, removing the need to pre-specify the number of critique iterations.
  arXiv  Detail & Related papers  (2025-04-21T04:56:47Z)
• Estimating Commonsense Plausibility through Semantic Shifts [66.06254418551737]
  We propose ComPaSS, a novel discriminative framework that quantifies commonsense plausibility by measuring semantic shifts.<n> Evaluations on two types of fine-grained commonsense plausibility estimation tasks show that ComPaSS consistently outperforms baselines.
  arXiv  Detail & Related papers  (2025-02-19T06:31:06Z)
• Evaluating Human Alignment and Model Faithfulness of LLM Rationale [66.75309523854476]
  We study how well large language models (LLMs) explain their generations through rationales. We show that prompting-based methods are less "faithful" than attribution-based explanations.
  arXiv  Detail & Related papers  (2024-06-28T20:06:30Z)
• Intent-conditioned and Non-toxic Counterspeech Generation using Multi-Task Instruction Tuning with RLAIF [14.2594830589926]
  Counterspeech, defined as a response to online hate speech, is increasingly used as a non-censorial solution. Our study introduces CoARL, a novel framework enhancing counterspeech generation by modeling the pragmatic implications underlying social biases in hateful statements. CoARL's first two phases involve sequential multi-instruction tuning, teaching the model to understand intents, reactions, and harms of offensive statements, and then learning task-specific low-rank adapter weights for generating intent-conditioned counterspeech.
  arXiv  Detail & Related papers  (2024-03-15T08:03:49Z)
• Toward Practical Automatic Speech Recognition and Post-Processing: a Call for Explainable Error Benchmark Guideline [12.197453599489963]
  We propose the development of an Error Explainable Benchmark (EEB) dataset. This dataset, while considering both speech- and text-level, enables a granular understanding of the model's shortcomings. Our proposition provides a structured pathway for a more real-world-centric' evaluation, allowing for the detection and rectification of nuanced system weaknesses.
  arXiv  Detail & Related papers  (2024-01-26T03:42:45Z)
• Measuring Fairness of Text Classifiers via Prediction Sensitivity [63.56554964580627]
  ACCUMULATED PREDICTION SENSITIVITY measures fairness in machine learning models based on the model's prediction sensitivity to perturbations in input features. We show that the metric can be theoretically linked with a specific notion of group fairness (statistical parity) and individual fairness.
  arXiv  Detail & Related papers  (2022-03-16T15:00:33Z)
• Characterizing the adversarial vulnerability of speech self-supervised learning [95.03389072594243]
  We make the first attempt to investigate the adversarial vulnerability of such paradigm under the attacks from both zero-knowledge adversaries and limited-knowledge adversaries. The experimental results illustrate that the paradigm proposed by SUPERB is seriously vulnerable to limited-knowledge adversaries.
  arXiv  Detail & Related papers  (2021-11-08T08:44:04Z)

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.