Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Compositional generalization through abstract representations in human and artificial neural networks
Takuya Ito, Tim Klinger, Doug H Schultz, John D Murray, Michael Cole, Mattia Rigotti
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 |
|---|---|---|---|
| add Primitives pretraining (ablation) | 2-layer MLP · C-PRO zero-shot | accuracy (%): 23.25% -> 31.51% | Improved |
| add Simple task pretraining (ablation) | 2-layer MLP · C-PRO zero-shot | accuracy (%): 23.25% -> 70.57% | Improved |
| combine Primitives and Simple task pretraining (ablation) | 2-layer MLP · C-PRO zero-shot | accuracy (%): 70.57% -> 92.15% | Improved |
| add Primitives pretraining (ablation) | 2-layer MLP · C-PRO novel contexts after 4 practiced | accuracy (%): 28.79% -> 57.97% | Improved |
| add Simple task pretraining (ablation) | 2-layer MLP · C-PRO novel contexts after 4 practiced | accuracy (%): 57.97% -> 86.79% | Improved |
| add Combined pretraining on top of Simple task pretraining (ablation) | 2-layer MLP · parallelism score | PS | No clear effect |
Closest papers by meaning
- CraftFactory: A Conditioned Control Policy Benchmark for Compositional Generalization (AAAI 2025 · similarity 0.88)
- When does compositional structure yield compositional generalization? A kernel theory. (ICLR 2025 · similarity 0.88)
- Scaling can lead to compositional generalization (NeurIPS 2025 · similarity 0.88)
- Evaluating alignment between humans and neural network representations in image-based learning tasks (NeurIPS 2024 · similarity 0.87)
- Separating the 'what' and 'how' of compositional computation to enable reuse and continual learning (NeurIPS 2025 · similarity 0.87)
- Swing-by Dynamics in Concept Learning and Compositional Generalization (ICLR 2025 · 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.
