CausalVQA

How to read this picture

Video question set on cause and effect in real Ego-Exo4D activity clips: counterfactual, anticipation, planning.1

Sources
Last checked 10 Oct 2026Basic entry15 of 22 facts checked at the sourceNext check 8 Apr 2027
Runs in
Recorded data1
Checked against real robots
Not checked
Skill
Reasoning
Robot
No body1
Used by
382
citations
Licence
Unclear3

Comparisons with real robots

No comparison found Unknown

Details

Not built to predict robot performance. Searched paper, repo.

Known problems 1

  1. Introduction and human-evaluation section say 1,786 items; composition section and datashe

    Internal inconsistency in arXiv v1; not resolved.1

Details

About

What it is
Benchmark Inferred4
More

Classified by the Atlas from how the authors describe and distribute it.

Built by
All six authors: FAIR at Meta1
Released
2025-06 (arXiv v1 2025-06-11)4
Version
arXiv v1 only4
Last update
Last repo commit 2025-08-18 'fixed quality of life functions for submission'.5
More

GitHub API.

Status
Maintained Inferred5

Setup

Runs in
Recorded data Inferred1
Robot
No body1
Setting
Mixed1
Size
1,586 items (793 paired questions), 779 video segments (Sec. 2.3.1 and datasheet).1

Scoring and access

Scored by
Accuracy1
Leaderboard
Official leaderboard6
Code licence
LICENSE.pdf in repo is the Ego-Exo4D licence agreement draft (licensor Carnegie Mellon University); no separate code licence. Model scripts adapted from lmms-eval.3
More

GitHub reports no licence.

Data licence
Benchmark available under the Ego-Exo4D licence; data fetched from the Ego4D consortium S3 bucket after sign-up and approval (up to 48 h).7
Access
By application Inferred7
Published at
Unknown
More

Searched OpenReview API (term 'CausalVQA') and Semantic Scholar (venue arXiv.org).

Sources 7

  1. 1CausalVQA: A Physically Grounded Causal Reasoning Benchmark for Video ModelsPaper · Jun 2025 · checked 10 Oct 2026
  2. 2Semantic Scholar API recordIndex · checked 10 Oct 2026
  3. 3facebookresearch/CausalVQA on GitHub (file LICENSE.pdf)Repository · checked 10 Oct 2026
  4. 4CausalVQA: A Physically Grounded Causal Reasoning Benchmark for Video ModelsPaper · Jun 2025 · checked 10 Oct 2026
  5. 5facebookresearch/CausalVQA on GitHub (repository)Repository · checked 10 Oct 2026
  6. 6facebook/physical_reasoning_leaderboard on Hugging Face (space)Leaderboard · checked 10 Oct 2026
  7. 7facebookresearch/CausalVQA on GitHub (file README.md)Repository · checked 10 Oct 2026

Change history

  1. Created as a basic entry: identity facts checked at primary sources (phase 1 re-verification).