# AGQA 2.0: An Updated Benchmark for Compositional Spatio-Temporal Reasoning

arXiv: https://arxiv.org/abs/2204.06105 · HTML: https://mlsanity.com/paper/2204.06105/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 5 of its 13 references that are in the map.
- First posted on arXiv 12 Apr 2022.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)

## Builds on (most cited first, 5 of 5)

- [Hierarchical Conditional Relation Networks for Video Question Answering](https://mlsanity.com/paper/2002.10698/) (2020)
- [REPAIR: Removing Representation Bias by Dataset Resampling](https://mlsanity.com/paper/1904.07911/) (2019)
- [Heterogeneous Memory Enhanced Multimodal Attention Model for Video Question Answering](https://mlsanity.com/paper/1904.04357/) (2019)
- [AGQA: A Benchmark for Compositional Spatio-Temporal Reasoning](https://mlsanity.com/paper/2103.16002/) (2021)
- [What Gives the Answer Away? Question Answering Bias Analysis on Video QA Datasets](https://mlsanity.com/paper/2007.03626/) (2020)

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

- [EgoSchema: A Diagnostic Benchmark for Very Long-form Video Language Understanding](https://mlsanity.com/paper/2308.09126/) (2023)
- [MIST: Multi-modal Iterative Spatial-Temporal Transformer for Long-form Video Question Answering](https://mlsanity.com/paper/2212.09522/) (2022)
- [Learning Situation Hyper-Graphs for Video Question Answering](https://mlsanity.com/paper/2304.08682/) (2023)
- [SViTT: Temporal Learning of Sparse Video-Text Transformers](https://mlsanity.com/paper/2304.08809/) (2023)
- [Neural-Symbolic VideoQA: Learning Compositional Spatio-Temporal Reasoning for Real-world Video Question Answering](https://mlsanity.com/paper/2404.04007/) (2024)
- [DyGEnc: Encoding a Sequence of Textual Scene Graphs to Reason and Answer Questions in Dynamic Scenes](https://mlsanity.com/paper/2505.03581/) (2025)
- [Learning to Reason Iteratively and Parallelly for Complex Visual Reasoning Scenarios](https://mlsanity.com/paper/2411.13754/) (2024)
- [Learning Fine-Grained Visual Understanding for Video Question Answering via Decoupling Spatial-Temporal Modeling](https://mlsanity.com/paper/2210.03941/) (2022)
- [POVQA: Preference-Optimized Video Question Answering with Rationales for Data Efficiency](https://mlsanity.com/paper/2510.01009/) (2025)
- [IntentQA: Intent Question Answering in Videos by Cognitive Context Reasoning](https://mlsanity.com/paper/2608.23330/) (2026)

## Abstract (arXiv)

Prior benchmarks have analyzed models' answers to questions about videos in order to measure visual compositional reasoning. Action Genome Question Answering (AGQA) is one such benchmark. AGQA provides a training/test split with balanced answer distributions to reduce the effect of linguistic biases. However, some biases remain in several AGQA categories. We introduce AGQA 2.0, a version of this benchmark with several improvements, most namely a stricter balancing procedure. We then report results on the updated benchmark for all experiments.