# ELV-Halluc: Benchmarking Semantic Aggregation Hallucinations in Long Video Understanding

arXiv: https://arxiv.org/abs/2508.21496 · HTML: https://mlsanity.com/paper/2508.21496/

## Key facts

- Cited by 10 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 23 of its 42 references that are in the map.
- First posted on arXiv 29 Aug 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, 23 of 23)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://mlsanity.com/paper/2104.09864/) (2021)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models](https://mlsanity.com/paper/2504.10479/) (2025)
- [LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models](https://mlsanity.com/paper/2403.13372/) (2024)
- [Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis](https://mlsanity.com/paper/2405.21075/) (2024)
- [MVBench: A Comprehensive Multi-modal Video Understanding Benchmark](https://mlsanity.com/paper/2311.17005/) (2023)
- [EgoSchema: A Diagnostic Benchmark for Very Long-form Video Language Understanding](https://mlsanity.com/paper/2308.09126/) (2023)
- [VLMEvalKit: An Open-Source Toolkit for Evaluating Large Multi-Modality Models](https://mlsanity.com/paper/2407.11691/) (2024)
- [LVBench: An Extreme Long Video Understanding Benchmark](https://mlsanity.com/paper/2406.08035/) (2024)
- [SWIFT:A Scalable lightWeight Infrastructure for Fine-Tuning](https://mlsanity.com/paper/2408.05517/) (2024)
- [SmolVLM: Redefining small and efficient multimodal models](https://mlsanity.com/paper/2504.05299/) (2025)
- [Video-MMMU: Evaluating Knowledge Acquisition from Multi-Discipline Professional Videos](https://mlsanity.com/paper/2501.13826/) (2025)
- [Video-Holmes: Can MLLM Think Like Holmes for Complex Video Reasoning?](https://mlsanity.com/paper/2505.21374/) (2025)
- [VideoHallucer: Evaluating Intrinsic and Extrinsic Hallucinations in Large Video-Language Models](https://mlsanity.com/paper/2406.16338/) (2024)
- [E.T. Bench: Towards Open-Ended Event-Level Video-Language Understanding](https://mlsanity.com/paper/2409.18111/) (2024)
- [VidHalluc: Evaluating Temporal Hallucinations in Multimodal Large Language Models for Video Understanding](https://mlsanity.com/paper/2412.03735/) (2024)
- [EventHallusion: Diagnosing Event Hallucinations in Video LLMs](https://mlsanity.com/paper/2409.16597/) (2024)
- [ARGUS: Hallucination and Omission Evaluation in Video-LLMs](https://mlsanity.com/paper/2506.07371/) (2025)

## Built on by (most cited first, 10 of 10)

- [EVA: Efficient Reinforcement Learning for End-to-End Video Agent](https://mlsanity.com/paper/2603.22918/) (2026)
- [VirtueBench: Evaluating Trustworthiness under Uncertainty in Long Video Understanding](https://mlsanity.com/paper/2603.07071/) (2026)
- [Learning to Decode Against Compositional Hallucination in Video Multimodal Large Language Models](https://mlsanity.com/paper/2602.00559/) (2026)
- [Stress Tests REVEAL Fragile Temporal and Visual Grounding in Video-Language Models](https://mlsanity.com/paper/2602.11244/) (2026)
- [Distorted or Fabricated? A Survey on Hallucination in Video LLMs](https://mlsanity.com/paper/2604.12944/) (2026)
- [Towards Temporal Compositional Reasoning in Long-Form Sports Videos](https://mlsanity.com/paper/2604.22226/) (2026)
- [Tracking the Truth: Object-Centric Spatio-Temporal Monitoring for Video Large Language Models](https://mlsanity.com/paper/2605.08974/) (2026)
- [Can We Trust Video Hallucination Detectors? VidHalLoc for Evaluating the Evaluators](https://mlsanity.com/paper/2609.09895/) (2026)
- [Beneath the Scores: Rethinking Hallucination Evaluation for Video Understanding Models](https://mlsanity.com/paper/2609.28991/) (2026)
- [STRAND: Benchmarking and Improving Object-Centric Spatio-Temporal Monitoring in Video Large Language Models](https://mlsanity.com/paper/2609.29607/) (2026)

## Abstract (arXiv)

Video multimodal large language models (Video-MLLMs) have achieved remarkable progress in video understanding. However, they remain vulnerable to hallucination-producing content inconsistent with or unrelated to video inputs. Previous video hallucination benchmarks primarily focus on short-videos. They attribute hallucinations to factors such as strong language priors, missing frames, or vision-language biases introduced by the visual encoder. While these causes indeed account for most hallucinations in short videos, they still oversimplify the cause of hallucinations. Sometimes, models generate incorrect outputs but with correct frame-level semantics. We refer to this type of hallucination as Semantic Aggregation Hallucination (SAH), which arises during the process of aggregating frame-level semantics into event-level semantic groups. Given that SAH becomes particularly critical in long videos due to increased semantic complexity across multiple events, it is essential to separate and thoroughly investigate the causes of this type of hallucination. To address the above issues, we introduce ELV-Halluc, the first benchmark dedicated to long-video hallucination, enabling a systematic investigation of SAH. Our experiments confirm the existence of SAH and show that it increases with semantic complexity. Additionally, we find that models are more prone to SAH on rapidly changing semantics. Moreover, we discuss potential approaches to mitigate SAH. We demonstrate that positional encoding strategy contributes to alleviating SAH, and further adopt DPO strategy to enhance the model's ability to distinguish semantics within and across events. To support this, we curate a dataset of 8K adversarial data pairs and achieve improvements on both ELV-Halluc and Video-MME, including a substantial 27.7% reduction in SAH ratio.