Research map / Probabilistic and scientific modeling / Causal discovery
Properties from mechanisms: an equivariance perspective on identifiable representation learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use the known latent transition mechanism to constrain the encoder | Identifiable representation learning with a known deterministic mechanism · Deterministic latent dynamics | identifiability | Helped |
| Use multiple known mechanisms rather than a single mechanism | Identifiable representation learning · Multiple deterministic mechanisms | identifiability | Helped |
| Allow a hypothesis class of possible mechanisms instead of assuming the mechanism is known | Identifiable representation learning · Unknown deterministic mechanisms | identifiability | Hurt |
| Use multiple affine mechanisms with changing offsets under the stated assumptions | Identifiable representation learning · Affine mechanisms | identifiability | Helped |
| Model latent evolution as stochastic mechanisms and match conditional observation distributions | Identifiable representation learning · Stochastic latent dynamics | identifiability | Helped |
| Interpret the generalized-Laplace transition assumptions as a stochastic mechanism | Existing identifiable representation learning result · Generalized-Laplace latent transitions | identifiability | Helped |
| Use sufficiently small non-Gaussian latent increments | Identifiable representation learning · Small-increment stochastic dynamics | identifiability | Helped |
Closest papers by meaning
- Identifiable Exchangeable Mechanisms for Causal Structure and Representation Learning (ICLR 2025 · similarity 0.90)
- Nonparametric Identification of Latent Concepts (ICML 2025 · similarity 0.88)
- Learning Linear Causal Representations from General Environments: Identifiability and Intrinsic Ambiguity (NeurIPS 2024 · similarity 0.87)
- Cross-Entropy Is All You Need To Invert the Data Generating Process (ICLR 2025 · similarity 0.87)
- Multi-View Causal Representation Learning with Partial Observability (ICLR 2024 · similarity 0.87)
- Concept Modulation Models: A Unified Framework for Identifiability and Extrapolation (NeurIPS 2026 · similarity 0.87)
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.