Datasets › HallusionBench
HallusionBench
Large language models (LLMs), after being aligned with vision models and integrated into vision-language models (VLMs), can bring impressive improvement in image reasoning tasks. This was shown by the recently released GPT-4V(ison), LLaVA-1.5, etc. However, the strong language prior in these SOTA LVLMs can be a double-edged sword: they may ignore the image context and solely rely on the (even contradictory) language prior for reasoning. In contrast, the vision modules in VLMs are weaker than LLMs and may result in misleading visual representations, which are then translated to confident mistakes by LLMs.
To study these two types of VLM mistakes, i.e., language hallucination and visual illusion, we curated HallusionBench, an image-context reasoning benchmark that is still challenging to even GPT-4V and LLaVA-1.5. We provide a detailed analysis of examples in HallusionBench, which sheds novel insights on the illusion or hallucination of VLMs and how to improve them in the future.
Benchmarks archive 2025-07-28
All 1 leaderboard whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.
Syntology 1 newer paper was judged to report on this dataset's benchmarks so far (Claude Sonnet 4.5, reading their tables: a model's judgement, not a result; 0 placed): Visual Question Answering (VQA) · HallusionBench (1 judged, 0 placed). Drawn from 8,886 of 18,637 newer papers checked so far (733 more had nothing to place; 41 could not be read).
| First row (archive order) | Paper | Code | ||||
|---|---|---|---|---|---|---|
| Visual Question Answering (VQA) | HallusionBench | GPT-4V Question Pair Acc 12.2047 | HallusionBench: An Advanced Diagnostic Suite for... | codelion/adaptive-classifier +8 | 4 | Compare |
Papers archive 2025-07-28
papers with a benchmark row: 3 · with a code link: 3 · where Syntology ran a sample: 1 (1 with a run with no instrument failure, 0 where every run was a failure of Syntology's instrument) Syntology
Show: all papers with a benchmark rowonly where code ran (1 of 3 with a benchmark row: 1 with a run with no instrument failure, 0 where every run was a failure of Syntology's instrument)
Page 1 of 1: papers 1 to 3 of 3 with a leaderboard row on this dataset's benchmarks, newest first by the archive's date (ties by slug; undated papers last).
The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset, not that list; the archive's count for this dataset is 53. The Syntology column is from Syntology's graph, stated per sample; it is not part of any archive number. A line reads “N ran (of which C constructed an object rather than computing a result; K with no instrument failure: H honoured, V violated, P with no contract checked; I where Syntology's instrument failed) · U unverified” (C is a part of N, never taken away from it); the figure “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. “Ran” means the sample executed on a synthesized fixture, not that the paper's result was reproduced. When the archive marks a repository official for the paper, the cell starts with that repository's state (the archive's flag, not a verdict on who wrote the code; “community repositories only” when every sample that ran came from a community repository, “official: no sample here; runs from other or unrecorded repositories” when some came from a repository the paper names or has in its text, or from none recorded); hover it for the repositories the samples that ran came from. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record.
| Paper | Code | Results | Date | Samples run Syntology |
|---|---|---|---|---|
| HallusionBench: An Advanced Diagnostic Suite for Entangled Language Hallucination and Visual Illusion in Large Vision-Language Models | 9 | 1 | 23 Oct 2023 | official (archive's flag): 1 ran · 4 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 1 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 4 unverified (1 pointer-only for licence) |
| Mitigating Hallucination in Large Multi-Modal Models via Robust Instruction Tuning | 4 | 1 | 26 Jun 2023 | not harvested |
| mPLUG-Owl: Modularization Empowers Large Language Models with Multimodality | 1 | 1 | 27 Apr 2023 | not harvested |
Dataset loaders archive 2025-07-28
1 loader as listed in the archive; links are outbound and not re-checked here.
Tasks archive 2025-07-28
License archive 2025-07-28
BSD 3-Clause License
Modalities archive 2025-07-28
Languages archive 2025-07-28
Variants archive 2025-07-28
- HallusionBench
1 variant name, as the archive lists them.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections