Research map / Probabilistic and scientific modeling / Causal discovery
Linear Causal Representation Learning by Topological Ordering, Pruning, and Disentanglement
Hao Chen, Lin Liu, Yu Guang Wang
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 changed | Where | Measured | Effect |
|---|---|---|---|
| use CREATOR (topological ordering, pruning, disentanglement) instead of LiNGCReL | linear causal representation learning · Synthetic, setting (1): noise distribution varies across environments, K=d, d=2,3,5,7 | LocR2 and SHD | Improved |
| use CREATOR instead of LiNGCReL when noise is invariant across environments but differs across components | linear causal representation learning · Synthetic, setting (2), K=d | LocR2 and SHD | Improved |
| use CREATOR instead of LiNGCReL to recover the BG, CD, ED causal DAG from LLM last hidden states | six LLMs (Llama Guard, Llama 3.1 Instruct, TinyLlama, Phi-3-Mini, GPT-Neo, BLOOM) · LLM story case study (K=3 story styles) | correct DAG edges | Mixed |
| weaken causal influence by scaling W by sigma from 0.005 to 0.5 so topological ordering inference has substantial error (ablation) | CREATOR · Synthetic, appendix E.2 | latent feature recovery | Hurt |
Closest papers by meaning
- Causal Representation Learning Made Identifiable by Grouping of Observational Variables (ICML 2024 · similarity 0.92)
- Learning Linear Causal Representations from General Environments: Identifiability and Intrinsic Ambiguity (NeurIPS 2024 · similarity 0.91)
- Identifying General Mechanism Shifts in Linear Causal Representations (NeurIPS 2024 · similarity 0.91)
- Unifying Causal Representation Learning with the Invariance Principle (ICLR 2025 · similarity 0.91)
- CausalVerse: Benchmarking Causal Representation Learning with Configurable High-Fidelity Simulations (NeurIPS 2025 · similarity 0.90)
- Sanity Checking Causal Representation Learning on a Simple Real-World System (ICML 2025 · similarity 0.90)
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.
