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

Source: [AGQA 2.0: An Updated Benchmark for Compositional Spatio-Temporal Reasoning](<https://www.sota2.com/research/paper/68fa6c3a1e209131cea32e83>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T00:06:48.603Z

Authors: Madeleine Grunde-McLaughlin, Ranjay Krishna, Maneesh Agrawala

Published: 2022-04-12T22:30:12.000Z

## Abstract

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.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2204.06105>)

[PDF](<https://arxiv.org/pdf/2204.06105.pdf>)