Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Homomorphism AutoEncoder --- Learning Group Structured Representations from Observed Transitions
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 HAE's group-structured latent dynamics instead of an unstructured latent forward model | HAE · dSprites 128-step rollout prediction | reconstruction loss | Helped |
| Use HAE's group-structured latent dynamics instead of an unstructured latent forward model | HAE · 3D bunny SO(3) 128-step rollout prediction | reconstruction error | Helped |
| Use a 2-step latent prediction loss rather than relying on 1-step latent prediction alone (ablation) | HAE · group-structured representation learning | group representation structure | Helped |
| Add an observation reconstruction loss to latent prediction (ablation) | HAE · group-structured representation learning | encoder non-collapse | Helped |
| Use 2-step reconstruction loss instead of reconstructing only the initial observation (ablation) | HAE · joint learning of the representation and group action | empirical performance | Helped |
| Add sparsity regularization on the learned group representation (ablation) | HAE · dSprites translations and bunny SO(3) | block-diagonal structure of the representation | Helped |
Closest papers by meaning
- Learning Group Actions on Latent Representations (NeurIPS 2024 · similarity 0.90)
- Disentangled representation learning through unsupervised symmetry group discovery (ICLR 2026 · similarity 0.88)
- Disentangling Representations through Multi-task Learning (ICLR 2025 · similarity 0.86)
- Act to See, See to Act: Diffusion-Driven Perception-Action Interplay for Adaptive Policies (NeurIPS 2025 · similarity 0.85)
- Equivariance by Contrast: Identifiable Equivariant Embeddings from Unlabeled Finite Group Actions (NeurIPS 2025 · similarity 0.85)
- Sim-to-Real Causal Transfer: A Metric Learning Approach to Causally-Aware Interaction Representations (CVPR 2025 · similarity 0.85)
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.