# Explanation-Driven Counterfactual Testing for Faithfulness in Vision-Language Model Explanations

arXiv: https://arxiv.org/abs/2510.00047 · HTML: https://mlsanity.com/paper/2510.00047/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Placed in VLM benchmarks and visual reasoning, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 17 of its 23 references that are in the map.
- First posted on arXiv 27 Sep 2025.

## Placed in

- Multimodal LLMs / vision-language models / [VLM benchmarks and visual reasoning](https://mlsanity.com/mllm/vlm-benchmarks-and-visual-reasoning/)

## Builds on (most cited first, 17 of 17)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization](https://mlsanity.com/paper/1610.02391/) (2016)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [Axiomatic Attribution for Deep Networks](https://mlsanity.com/paper/1703.01365/) (2017)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations](https://mlsanity.com/paper/2108.01073/) (2021)
- [A Survey on LLM-as-a-Judge](https://mlsanity.com/paper/2411.15594/) (2024)
- [Sanity Checks for Saliency Maps](https://mlsanity.com/paper/1810.03292/) (2018)
- [OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge](https://mlsanity.com/paper/1906.00067/) (2019)
- [Towards Faithfully Interpretable NLP Systems: How should we define and evaluate faithfulness?](https://mlsanity.com/paper/2004.03685/) (2020)
- [LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methods](https://mlsanity.com/paper/2412.05579/) (2024)
- [Counterfactual Visual Explanations](https://mlsanity.com/paper/1904.07451/) (2019)
- [OmniGen2: Towards Instruction-Aligned Multimodal Generation](https://mlsanity.com/paper/2506.18871/) (2025)
- [Counterfactual Explanations Can Be Manipulated](https://mlsanity.com/paper/2106.02666/) (2021)
- [Faithful Multimodal Explanation for Visual Question Answering](https://mlsanity.com/paper/1809.02805/) (2018)
- [CounterCurate: Enhancing Physical and Semantic Visio-Linguistic Compositional Reasoning via Counterfactual Examples](https://mlsanity.com/paper/2402.13254/) (2024)

## Built on by (most cited first, 12 of 12)

- [On Robustness and Chain-of-Thought Consistency of RL-Finetuned VLMs](https://mlsanity.com/paper/2602.12506/) (2026)
- [When to Call an Apple Red: Humans Follow Introspective Rules, VLMs Don't](https://mlsanity.com/paper/2604.06422/) (2026)
- [Position, Not Provenance: Separating Reasoning Mediation from Sycophancy in Medical Vision-Language Models](https://mlsanity.com/paper/2607.27304/) (2026)
- [BEVLM: Distilling Semantic Knowledge from LLMs into Bird's-Eye View Representations](https://mlsanity.com/paper/2603.06576/) (2026)
- [Revealing the Gap in Human and VLM Scene Perception through Counterfactual Semantic Saliency](https://mlsanity.com/paper/2605.13047/) (2026)
- [Faithful Action-unit Causal Reasoning for Counterfactually Faithful Emotion Explanations](https://mlsanity.com/paper/2606.15779/) (2026)
- [When Do Frozen VLMs Respond to Image-Free Object-Token Edits? An Answer-Key-Free Protocol and What It Reveals](https://mlsanity.com/paper/2609.03429/) (2026)
- [Beyond the Verdict: Evidence-Aligned Evaluation of Visual Prompt-Injection Guardrails](https://mlsanity.com/paper/2609.05535/) (2026)
- [Counterfactual Tests for Measuring Chain-of-Thought Faithfulness in Visual Language Models](https://mlsanity.com/paper/2609.06704/) (2026)
- [Does a model's stated reason for rejecting a candidate do any work?](https://mlsanity.com/paper/2609.30151/) (2026)
- [AnswerMap: Faithful Spatial Interpretability of VLMs from Answer Posteriors](https://mlsanity.com/paper/2609.35247/) (2026)
- [Visual sensitivity is not claim retractability: persistence-aware credit assignment for multimodal reinforcement learning](https://mlsanity.com/paper/2609.36572/) (2026)

## Abstract (arXiv)

Vision-Language Models (VLMs) often produce fluent Natural Language Explanations (NLEs) that sound convincing but may not reflect the causal factors driving predictions. This mismatch of plausibility and faithfulness poses technical and governance risks. We introduce Explanation-Driven Counterfactual Testing (EDCT), a fully automated verification procedure for a target VLM that treats the model's own explanation as a falsifiable hypothesis. Given an image-question pair, EDCT: (1) obtains the model's answer and NLE, (2) parses the NLE into testable visual concepts, (3) generates targeted counterfactual edits via generative inpainting, and (4) computes a Counterfactual Consistency Score (CCS) using LLM-assisted analysis of changes in both answers and explanations. Across 120 curated OK-VQA examples and multiple VLMs, EDCT uncovers substantial faithfulness gaps and provides regulator-aligned audit artifacts indicating when cited concepts fail causal tests.