Research map / Vision-language and multimodal / Multimodal LLMs
HaloQuest: A Visual Hallucination Dataset for Advancing Multimodal Reasoning
Zhecan Wang, Garrett Bingham, Adams Wei Yu, Quoc V. Le, Thang Luong, Golnaz Ghiasi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| fine-tune BLIP2 (8B) on VQA v2 plus HaloQuest instead of VQA v2 only (ablation) | BLIP2 (8B) · HaloQuest eval | accuracy (Auto-Eval): 12.3 -> 34.7 | Improved |
| fine-tune MiniGPT4 (13B) on VQA v2 plus HaloQuest (ablation) | MiniGPT4 (13B) · HaloQuest eval | accuracy (Auto-Eval): 16.1 -> 40.1 | Improved |
| add HaloQuest training data (ablation) | BLIP2 (8B) · VQA v2 (overall) | exact match accuracy: 70.5 -> 71.0 | Improved |
| add HaloQuest training data (ablation) | MiniGPT4 (13B) · VQA v2 (overall) | exact match accuracy: 74.9 -> 74.9 | No clear effect |
| train MiniGPT4 on HaloQuest | MiniGPT4 · POPE Random | accuracy: 0.72 -> 0.80 | Improved |
| use advanced Langfun schema with main-point reasoning instead of text-only prompt (ablation) | Auto-Eval · Auto-Eval agreement with human raters | agreement (%): 93.4 -> 95.3 | Improved |
| scale model size | zero-shot VLMs · HaloQuest eval (zero-shot) | accuracy (Auto-Eval): BEiT-3 (0.7B) 40.0 vs LLaVA (13B) 10.9 | No clear effect |
| evaluate on generated images instead of real images | benchmark design · HaloQuest eval | correlation across models (r): r=0.97 | No clear effect |
| use Auto-Eval instead of BLEU, CIDEr, ROUGE, METEOR as hallucination metric | evaluation metric · HaloQuest eval | correlation with human evaluation (r): r=0.99 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MiniGPT4 — MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models (ICLR 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- PhD: A ChatGPT-Prompted Visual Hallucination Evaluation Dataset (CVPR 2025 · similarity 0.93)
- Fine-Grained Multi Image Object Hallucination Benchmark (CVPR 2026 · similarity 0.93)
- MIRAGE: Assessing Hallucination in Multimodal Reasoning Chains of MLLM (NeurIPS 2025 · similarity 0.93)
- Detecting and Preventing Hallucinations in Large Vision Language Models (AAAI 2024 · similarity 0.93)
- Analyzing and Mitigating Object Hallucination: A Training Bias Perspective (AAAI 2026 · similarity 0.92)
- Combating Multimodal LLM Hallucination via Bottom-Up Holistic Reasoning (AAAI 2025 · similarity 0.92)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
