{"id":12,"identifier":"LOR-00012","url":"https://lifeofresearch.com/articles/12","title":"Causal Representation Learning from Observational Video: Identifiability Assumptions and the Gap Between Synthetic and Real-World Settings","abstract":"Causal representation learning (CRL) — the recovery of latent causal variables and their causal graph from high-dimensional\nobservations — has produced a substantial body of theoretical results in the past five years. CITRIS, CauCA, CRID, mechanismsparsity methods, and the recent grouping-based identifiability framework all establish conditions under which causal\nstructure can in principle be recovered. The gap between this theoretical progress and what has actually been demonstrated\non real-world observational video is wider than most accounts acknowledge. Almost all empirical CRL work uses synthetic 3Drendered video sequences with cleanly designed interventions, ground-truth latent factors, and assumption-friendly\ngenerative processes. Real video provides none of these. This paper argues that the question is not whether CRL is\ntheoretically possible — it is, under specific assumptions — but which of those assumptions real observational video can\nplausibly satisfy and which it cannot. We make three claims. First, the identifiability literature implicitly assumes intervention\naccess, multi-environment data, or strong sparsity priors that real video provides only partially and noisily. Second, three\nproperties of real video — object near-independence, occlusions and scene changes as quasi-interventions, and naturally\nvarying environmental contexts — offer plausible but unverified routes to satisfying identifiability assumptions. Third, objectcentric video models (Slot Attention, SAVi, SlotFormer) provide a necessary inductive bias but do not by themselves recover\ncausal structure; the bridge between object-centric representation and causal identification is the genuine empirical frontier.\nWe propose a research agenda focused on this bridge, on benchmarks that span synthetic-to-real, and on the methodological\nhonesty needed to make cumulative progress.","authors":[{"name":"Pranay Mahendrakar","affiliation":"","orcid":"0009-0003-7224-029X"}],"type":"article","license":"CC-BY-4.0","keywords":[],"publication_date":"2026-04-28","version":1,"views":6,"downloads":5,"created_at":"2026-09-22 17:33:54","doi":"10.5281/zenodo.19854728","source":"Zenodo","source_url":"https://zenodo.org/records/19854728","files":[{"id":12,"name":"Causal_Representation_Learning_Video_Mahendrakar.pdf","size":93091,"mime":"application/pdf","downloads":5,"download_url":"https://lifeofresearch.com/articles/12/files/12/Causal_Representation_Learning_Video_Mahendrakar.pdf"}]}