AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

PACS: A Dataset for Physical Audiovisual Commonsense Reasoning

Samuel Yu, Peter Wu, Paul Pu Liang, Ruslan Salakhutdinov, Louis-Philippe Morency

ECCV 2022 · Vision-language and multimodal › Multimodal LLMs · approach: Chain-of-thought multimodal reasoning · 18 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add audio input for human annotators (ablation)Human · PACS testaccuracy (%): 90.5 -> 96.3Improved
add audio input (ablation)Merlot Reserve (Large) · PACS testaccuracy (%): 68.4 -> 70.1Improved
add audio input (ablation)Merlot Reserve (Base) · PACS testaccuracy (%): 64.0 -> 66.5Improved
add audio input (ablation)CLIP/AudioCLIP · PACS testaccuracy (%): 56.3 -> 60.0Improved
add audio input (ablation)Late Fusion · PACS testaccuracy (%): 52.5 -> 55.0Improved
use only audio plus question instead of all modalities (ablation)late-fusion bias model · PACS testaccuracy (%): 50.9 -> 55.0No clear effect
scale Merlot Reserve from Base to LargeMerlot Reserve · PACS test (with audio)accuracy (%): 66.5 -> 70.1Improved
change task from physical commonsense QA to material classificationAudioCLIP · PACS testaccuracy (%): 60.0 -> 75.9Improved
change task from physical commonsense QA to material classificationLate Fusion · PACS testaccuracy (%): 55.0 -> 67.4Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy