# Argument-Based Consistency in Toxicity Explanations of LLMs

arXiv: https://arxiv.org/abs/2506.19113 · HTML: https://mlsanity.com/paper/2506.19113/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 48 of its 101 references that are in the map.
- First posted on arXiv 23 Jun 2025.

## Builds on (most cited first, 30 of 48)

- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017)
- [Axiomatic Attribution for Deep Networks](https://mlsanity.com/paper/1703.01365/) (2017)
- [RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models](https://mlsanity.com/paper/2009.11462/) (2020)
- [Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting](https://mlsanity.com/paper/2305.04388/) (2023)
- [Towards Faithfully Interpretable NLP Systems: How should we define and evaluate faithfulness?](https://mlsanity.com/paper/2004.03685/) (2020)
- [ERASER: A Benchmark to Evaluate Rationalized NLP Models](https://mlsanity.com/paper/1911.03429/) (2019)
- [How Can We Know When Language Models Know? On the Calibration of Language Models for Question Answering](https://mlsanity.com/paper/2012.00955/) (2020)
- [HateXplain: A Benchmark Dataset for Explainable Hate Speech Detection](https://mlsanity.com/paper/2012.10289/) (2020)
- [Nuanced Metrics for Measuring Unintended Bias with Real Data for Text Classification](https://mlsanity.com/paper/1903.04561/) (2019)
- [The (Un)reliability of saliency methods](https://mlsanity.com/paper/1711.00867/) (2017)
- [Annotators with Attitudes: How Annotator Beliefs And Identities Bias Toxic Language Detection](https://mlsanity.com/paper/2111.07997/) (2021)
- [On Second Thought, Let's Not Think Step by Step! Bias and Toxicity in Zero-Shot Reasoning](https://mlsanity.com/paper/2212.08061/) (2022)
- [The Unreliability of Explanations in Few-shot Prompting for Textual Reasoning](https://mlsanity.com/paper/2205.03401/) (2022)
- [Fact-Checking the Output of Large Language Models via Token-Level Uncertainty Quantification](https://mlsanity.com/paper/2403.04696/) (2024)
- [Measuring Association Between Labels and Free-Text Rationales](https://mlsanity.com/paper/2010.12762/) (2020)
- [Towards Faithful Model Explanation in NLP: A Survey](https://mlsanity.com/paper/2209.11326/) (2022)
- [Question Decomposition Improves the Faithfulness of Model-Generated Reasoning](https://mlsanity.com/paper/2307.11768/) (2023)
- [Faithfulness Tests for Natural Language Explanations](https://mlsanity.com/paper/2305.18029/) (2023)
- [Faithfulness vs. Plausibility: On the (Un)Reliability of Explanations from Large Language Models](https://mlsanity.com/paper/2402.04614/) (2024)
- [Chain of Thoughtlessness? An Analysis of CoT in Planning](https://mlsanity.com/paper/2405.04776/) (2024)
- [DICES Dataset: Diversity in Conversational AI Evaluation for Safety](https://mlsanity.com/paper/2306.11247/) (2023)
- [Do Models Explain Themselves? Counterfactual Simulatability of Natural Language Explanations](https://mlsanity.com/paper/2307.08678/) (2023)
- [Aligning Faithful Interpretations with their Social Attribution](https://mlsanity.com/paper/2006.01067/) (2020)
- [ReCEval: Evaluating Reasoning Chains via Correctness and Informativeness](https://mlsanity.com/paper/2304.10703/) (2023)
- [Is Your Toxicity My Toxicity? Exploring the Impact of Rater Identity on Toxicity Annotation](https://mlsanity.com/paper/2205.00501/) (2022)
- [ConU: Conformal Uncertainty in Large Language Models with Correctness Coverage Guarantees](https://mlsanity.com/paper/2407.00499/) (2024)
- [Framework for Evaluating Faithfulness of Local Explanations](https://mlsanity.com/paper/2202.00734/) (2022)
- [Don't Trust: Verify -- Grounding LLM Quantitative Reasoning with Autoformalization](https://mlsanity.com/paper/2403.18120/) (2024)

## Built on by (most cited first, 2 of 2)

- [Reasoning About Reasoning: Towards Informed and Reflective Use of LLM Reasoning in HCI](https://mlsanity.com/paper/2510.22978/) (2025)
- [Evaluating Second-Order Bias of LLMs Through Epistemic Entitlement](https://mlsanity.com/paper/2606.17506/) (2026)

## Abstract (arXiv)

The discourse around toxicity and LLMs in NLP largely revolves around detection tasks. This work shifts the focus to evaluating LLMs' reasoning about toxicity - from their explanations that justify a stance - to enhance their trustworthiness in downstream tasks. Despite extensive research on explainability, it is not straightforward to adopt existing methods to evaluate free-form toxicity explanation due to their over-reliance on input text perturbations, among other challenges. To account for these, we propose a novel, theoretically-grounded multi-dimensional criterion, Argument-based Consistency (ArC), that measures the extent to which LLMs' free-form toxicity explanations reflect an ideal and logical argumentation process. Based on uncertainty quantification, we develop six metrics for ArC to comprehensively evaluate the (in)consistencies in LLMs' toxicity explanations. We conduct several experiments on three Llama models (of size up to 70B) and an 8B Ministral model on five diverse toxicity datasets. Our results show that while LLMs generate plausible explanations to simple prompts, their reasoning about toxicity breaks down when prompted about the nuanced relations between the complete set of reasons, the individual reasons, and their toxicity stances, resulting in inconsistent and irrelevant responses. We open-source our code (https://github.com/uofthcdslab/ArC) and LLM-generated explanations (https://huggingface.co/collections/uofthcdslab/arc) for future works.