# Treble Counterfactual VLMs: A Causal Approach to Hallucination

arXiv: https://arxiv.org/abs/2503.06169 · HTML: https://mlsanity.com/paper/2503.06169/

## Key facts

- Cited by 35 papers in the map (data through 14 Oct 2026).
- Placed in Hallucination mitigation in LVLMs, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 37 of its 55 references that are in the map.
- First posted on arXiv 8 Mar 2025.

## Placed in

- Multimodal LLMs / vision-language models / [Hallucination mitigation in LVLMs](https://mlsanity.com/mllm/hallucination-mitigation-in-lvlms/)

## Builds on (most cited first, 30 of 37)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Flamingo: a Visual Language Model for Few-Shot Learning](https://mlsanity.com/paper/2204.14198/) (2022)
- [InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning](https://mlsanity.com/paper/2305.06500/) (2023)
- [MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models](https://mlsanity.com/paper/2304.10592/) (2023)
- [A Survey of Large Language Models](https://mlsanity.com/paper/2303.18223/) (2023)
- [Survey of Hallucination in Natural Language Generation](https://mlsanity.com/paper/2202.03629/) (2022)
- [Evaluating Object Hallucination in Large Vision-Language Models](https://mlsanity.com/paper/2305.10355/) (2023)
- [Scaling Vision Transformers](https://mlsanity.com/paper/2106.04560/) (2021)
- [A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity](https://mlsanity.com/paper/2302.04023/) (2023)
- [Aligning Large Multimodal Models with Factually Augmented RLHF](https://mlsanity.com/paper/2309.14525/) (2023)
- [A-OKVQA: A Benchmark for Visual Question Answering using World Knowledge](https://mlsanity.com/paper/2206.01718/) (2022)
- [Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decoding](https://mlsanity.com/paper/2311.16922/) (2023)
- [Object Hallucination in Image Captioning](https://mlsanity.com/paper/1809.02156/) (2018)
- [ClipCap: CLIP Prefix for Image Captioning](https://mlsanity.com/paper/2111.09734/) (2021)
- [End-to-end Autonomous Driving: Challenges and Frontiers](https://mlsanity.com/paper/2306.16927/) (2023)
- [OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocation](https://mlsanity.com/paper/2311.17911/) (2023)
- [Analyzing and Mitigating Object Hallucination in Large Vision-Language Models](https://mlsanity.com/paper/2310.00754/) (2023)
- [Woodpecker: Hallucination Correction for Multimodal Large Language Models](https://mlsanity.com/paper/2310.16045/) (2023)
- [Multi-Modal Hallucination Control by Visual Information Grounding](https://mlsanity.com/paper/2403.14003/) (2024)
- [Interpreting and Editing Vision-Language Representations to Mitigate Hallucinations](https://mlsanity.com/paper/2410.02762/) (2024)
- [VPGTrans: Transfer Visual Prompt Generator across LLMs](https://mlsanity.com/paper/2305.01278/) (2023)
- [The Factual Inconsistency Problem in Abstractive Text Summarization: A Survey](https://mlsanity.com/paper/2104.14839/) (2021)
- [Invariant Grounding for Video Question Answering](https://mlsanity.com/paper/2206.02349/) (2022)
- [Multi-Object Hallucination in Vision-Language Models](https://mlsanity.com/paper/2407.06192/) (2024)
- [Reducing Hallucinations in Vision-Language Models via Latent Space Steering](https://mlsanity.com/paper/2410.15778/) (2024)
- [SCOUT: Self-aware Discriminant Counterfactual Explanations](https://mlsanity.com/paper/2004.07769/) (2020)

## Built on by (most cited first, 30 of 35)

- [LLM-based Agents Suffer from Hallucinations: A Survey of Taxonomy, Methods, and Directions](https://mlsanity.com/paper/2509.18970/) (2025)
- [Selftok: Discrete Visual Tokens of Autoregression, by Diffusion, and for Reasoning](https://mlsanity.com/paper/2505.07538/) (2025)
- [A Survey of Multimodal Hallucination Evaluation and Detection](https://mlsanity.com/paper/2507.19024/) (2025)
- [Exploring Hallucination of Large Multimodal Models in Video Understanding: Benchmark, Analysis and Mitigation](https://mlsanity.com/paper/2503.19622/) (2025)
- [FORTIS: Benchmarking Over-Privilege in Agent Skills](https://mlsanity.com/paper/2605.09163/) (2026)
- [Charts Are Not Images: On the Challenges of Scientific Chart Editing](https://mlsanity.com/paper/2512.00752/) (2025)
- [Defenses Against Prompt Attacks Learn Surface Heuristics](https://mlsanity.com/paper/2601.07185/) (2026)
- [Counterfactual Trace Auditing of LLM Agent Skills](https://mlsanity.com/paper/2605.11946/) (2026)
- [Tell Model Where to Look: Mitigating Hallucinations in MLLMs by Vision-Guided Attention](https://mlsanity.com/paper/2511.20032/) (2025)
- [The Autonomy Tax: Defense Training Breaks LLM Agents](https://mlsanity.com/paper/2603.19423/) (2026)
- [RaMem: Contextual Reinstatement for Long-term Agentic Memory](https://mlsanity.com/paper/2606.22844/) (2026)
- [M3OOD: Automatic Selection of Multimodal OOD Detectors](https://mlsanity.com/paper/2508.11936/) (2025)
- [FESTA: Functionally Equivalent Sampling for Trust Assessment of Multimodal LLMs](https://mlsanity.com/paper/2509.16648/) (2025)
- [MambaTAD: When State-Space Models Meet Long-Range Temporal Action Detection](https://mlsanity.com/paper/2511.17929/) (2025)
- [SAVeS: Steering Safety Judgments in Vision-Language Models via Semantic Cues](https://mlsanity.com/paper/2603.19092/) (2026)
- [Beyond the Embedding Bottleneck: Adaptive Retrieval-Augmented 3D CT Report Generation](https://mlsanity.com/paper/2603.15822/) (2026)
- [Memory Retrieval for Changing Preferences](https://mlsanity.com/paper/2606.02976/) (2026)
- [Inside the Unfair Judge: A Mechanistic Interpretability Account of LLM-as-Judge Bias](https://mlsanity.com/paper/2607.11871/) (2026)
- [Evidence-RL: Towards Evidence-intensive Visual Reasoning](https://mlsanity.com/paper/2608.08021/) (2026)
- [Once Correct, Still Wrong: Counterfactual Hallucination in Multilingual Vision-Language Models](https://mlsanity.com/paper/2602.05437/) (2026)
- [Geometry over Density: Few-Shot Cross-Domain OOD Detection](https://mlsanity.com/paper/2605.03410/) (2026)
- [Said Aloud, Read Different: Cross-Modal Instability in Multimodal Models](https://mlsanity.com/paper/2608.27135/) (2026)
- [HomeSafe-Bench: Evaluating Vision-Language Models on Unsafe Action Detection for Embodied Agents in Household Scenarios](https://mlsanity.com/paper/2603.11975/) (2026)
- [JigShape: Evaluating Visual-Geometric Reasoning in VLMs through Jigsaw Puzzles](https://mlsanity.com/paper/2607.27670/) (2026)
- [Strategic Fusion of Vision Language Models: Shapley-Credited Context-Aware Dawid-Skene for Multi-Label Tasks in Autonomous Driving](https://mlsanity.com/paper/2510.01126/) (2025)
- [System-Mediated Attention Imbalances Make Vision-Language Models Say Yes](https://mlsanity.com/paper/2601.12430/) (2026)
- [How Many Counterfactuals Does It Take? Probing VLM Hallucinations Through Circuits and Causal Effects](https://mlsanity.com/paper/2606.08777/) (2026)
- [ProWAFT: A ROMA-LPD Instance for Workload-Aware and Dynamic Fault Tolerance in FPGA-Based CNN Accelerators](https://mlsanity.com/paper/2607.01602/) (2026)
- [CertVLA: Certified Defense against Physical Visual Attacks for Vision-Language-Action Models](https://mlsanity.com/paper/2608.20791/) (2026)
- [LegendBench: A Diagnostic Benchmark for Legend Understanding with Counterfactual Interventions](https://mlsanity.com/paper/2609.24172/) (2026)

## Abstract (arXiv)

Vision-Language Models (VLMs) have advanced multi-modal tasks like image captioning, visual question answering, and reasoning. However, they often generate hallucinated outputs inconsistent with the visual context or prompt, limiting reliability in critical applications like autonomous driving and medical imaging. Existing studies link hallucination to statistical biases, language priors, and biased feature learning but lack a structured causal understanding. In this work, we introduce a causal perspective to analyze and mitigate hallucination in VLMs. We hypothesize that hallucination arises from unintended direct influences of either the vision or text modality, bypassing proper multi-modal fusion. To address this, we construct a causal graph for VLMs and employ counterfactual analysis to estimate the Natural Direct Effect (NDE) of vision, text, and their cross-modal interaction on the output. We systematically identify and mitigate these unintended direct effects to ensure that responses are primarily driven by genuine multi-modal fusion. Our approach consists of three steps: (1) designing structural causal graphs to distinguish correct fusion pathways from spurious modality shortcuts, (2) estimating modality-specific and cross-modal NDE using perturbed image representations, hallucinated text embeddings, and degraded visual inputs, and (3) implementing a test-time intervention module to dynamically adjust the model's dependence on each modality. Experimental results demonstrate that our method significantly reduces hallucination while preserving task performance, providing a robust and interpretable framework for improving VLM reliability. To enhance accessibility and reproducibility, our code is publicly available at https://github.com/TREE985/Treble-Counterfactual-VLMs.