# NeuS-QA: Grounding Long-Form Video Understanding in Temporal Logic and Neuro-Symbolic Reasoning

arXiv: https://arxiv.org/abs/2509.18041 · HTML: https://mlsanity.com/paper/2509.18041/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 18 of its 46 references that are in the map.
- First posted on arXiv 22 Sep 2025.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)
- Multimodal LLMs / vision-language models / [Video understanding with MLLMs](https://mlsanity.com/mllm/video-understanding-with-mllms/)

## Builds on (most cited first, 18 of 18)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [Qwen2 Technical Report](https://mlsanity.com/paper/2407.10671/) (2024)
- [LVBench: An Extreme Long Video Understanding Benchmark](https://mlsanity.com/paper/2406.08035/) (2024)
- [LMMs-Eval: Reality Check on the Evaluation of Large Multimodal Models](https://mlsanity.com/paper/2407.12772/) (2024)
- [Aria: An Open Multimodal Native Mixture-of-Experts Model](https://mlsanity.com/paper/2410.05993/) (2024)
- [CinePile: A Long Video Question Answering Dataset and Benchmark](https://mlsanity.com/paper/2405.08813/) (2024)
- [MovieChat+: Question-aware Sparse Memory for Long Video Question Answering](https://mlsanity.com/paper/2404.17176/) (2024)
- [Too Many Frames, Not All Useful: Efficient Strategies for Long-Form Video QA](https://mlsanity.com/paper/2406.09396/) (2024)
- [The Probabilistic Model Checker Storm](https://mlsanity.com/paper/2002.07080/) (2020)
- [Zero-Shot Video Question Answering with Procedural Programs](https://mlsanity.com/paper/2312.00937/) (2023)
- [ALLVB: All-in-One Long Video Understanding Benchmark](https://mlsanity.com/paper/2503.07298/) (2025)
- [Specification-Driven Video Search via Foundation Models and Formal Verification](https://mlsanity.com/paper/2309.10171/) (2023)
- [NeSyA: Neurosymbolic Automata](https://mlsanity.com/paper/2412.07331/) (2024)
- [We'll Fix it in Post: Improving Text-to-Video Generation with Neuro-Symbolic Feedback](https://mlsanity.com/paper/2504.17180/) (2025)
- [NeuroStrata: Harnessing Neurosymbolic Paradigms for Improved Design, Testability, and Verifiability of Autonomous CPS](https://mlsanity.com/paper/2502.12267/) (2025)
- [A Challenge to Build Neuro-Symbolic Video Agents](https://mlsanity.com/paper/2505.13851/) (2025)

## Built on by (most cited first, 8 of 8)

- [HiMu: Hierarchical Multimodal Frame Selection for Long Video Question Answering](https://mlsanity.com/paper/2603.18558/) (2026)
- [RT-NeuS: Towards Real-Time Neuro-Symbolic Video Understanding via Adaptive Temporal Verification](https://mlsanity.com/paper/2602.23553/) (2026)
- [Incentivizing Vision Language Models to Search for Long Video Question Answering](https://mlsanity.com/paper/2607.02959/) (2026)
- [Driving Video Retrieval for Complex Queries with Structured Grounding](https://mlsanity.com/paper/2606.09109/) (2026)
- [CrossView: Can Vision-Language Models Reason Across Cameras?](https://mlsanity.com/paper/2608.15539/) (2026)
- [Logi-PAR: Logic-Infused Patient Activity Recognition via Differentiable Rule](https://mlsanity.com/paper/2603.05184/) (2026)
- [TLG: Temporal-Logic Grounding for Video Question Answering via Source-Annotation Reconstruction and Category-Targeted Reasoning](https://mlsanity.com/paper/2606.01591/) (2026)
- [TL;QR: Querying Temporal Discursive Patterns in Textual Corpora](https://mlsanity.com/paper/2610.15712/) (2026)

## Abstract (arXiv)

While vision-language models (VLMs) excel at tasks involving single images or short videos, they still struggle with Long Video Question Answering (LVQA) due to its demand for complex multi-step temporal reasoning. Vanilla approaches, which simply sample frames uniformly and feed them to a VLM along with the question, incur significant token overhead. This forces aggressive downsampling of long videos, causing models to miss fine-grained visual structure, subtle event transitions, and key temporal cues. Recent works attempt to overcome these limitations through heuristic approaches; however, they lack explicit mechanisms for encoding temporal relationships and fail to provide any formal guarantees that the sampled context actually encodes the compositional or causal logic required by the question. To address these foundational gaps, we introduce NeuS-QA, a training-free, plug-and-play neuro-symbolic pipeline for LVQA. NeuS-QA first translates a natural language question into a logic specification that models the temporal relationship between frame-level events. Next, we construct a video automaton to model the video's frame-by-frame event progression, and finally employ model checking to compare the automaton against the specification to identify all video segments that satisfy the question's logical requirements. Only these logic-verified segments are submitted to the VLM, thus improving interpretability, reducing hallucinations, and enabling compositional reasoning without modifying or fine-tuning the model. Experiments on the LongVideoBench and CinePile LVQA benchmarks show that NeuS-QA significantly improves performance by over 10%, particularly on questions involving event ordering, causality, and multi-step reasoning. We open-source our code at https://utaustin-swarmlab.github.io/NeuS-QA/.