Research map / Label-efficient and robust learning / Recognition and long-tail
Formalizing Consistency and Coherence of Representation Learning
Harald Stromfelt, Luke Dickens, Artur Garcez, Alessandra Russo
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 Dynamic Comparator relation-decoder instead of HolE | relation-decoder · PRT MNIST->BlockStacks | aggregate coherence (consistency loss): 12.12 -> 0.53 | Improved |
| use Dynamic Comparator instead of NN | relation-decoder · PRT MNIST->BlockStacks | aggregate coherence (consistency loss): 1.71 -> 0.53 | Improved |
| use Dynamic Comparator instead of TransR | relation-decoder · PRT MNIST->BlockStacks | aggregate coherence (consistency loss): 2.51 -> 0.53 | Improved |
| use Dynamic Comparator instead of NTN | relation-decoder · PRT MNIST->BlockStacks | Con-A coherence (consistency loss): 1.92 -> 0.29 | Improved |
| use Dynamic Comparator instead of NN on reflexivity | relation-decoder · PRT MNIST->BlockStacks | Con-I-R coherence: 0.05 -> 0.09 | Mixed |
| increase disentanglement pressure beta in {1,4,8,12} (ablation) | DC · PRT MNIST->BlockStacks | target-domain accuracy of held-out relations | No clear effect |
| increase beta (ablation) | NN, NTN, HolE · PRT MNIST->BlockStacks | target-domain PRT accuracy | Mixed |
| increase beta (ablation) | TransR · PRT MNIST->BlockStacks | target-domain PRT accuracy | Improved |
| increase beta (ablation) | all models · PRT MNIST->BlockStacks | source-domain accuracy | No clear effect |
| transfer only the isSuccessor decoder to the target domain | NN, NTN, HolE · PRT MNIST->BlockStacks | accuracy on held-out relations | Hurt |
| evaluate consistency on interpolated latent samples instead of data embeddings | consistency losses · PRT MNIST->BlockStacks | Spearman correlation with PRT accuracy | Improved |
| increase beta (ablation) | NN, NTN · PRT MNIST->BlockStacks | Con-I consistency losses | Improved |
Closest papers by meaning
- Logically Consistent Language Models via Neuro-Symbolic Integration (ICLR 2025 · similarity 0.86)
- Deciphering Raw Data in Neuro-Symbolic Learning with Provable Guarantees (AAAI 2024 · similarity 0.86)
- Concise and Logically Consistent Conformal Sets for Neuro-Symbolic Concept-Based Models (NeurIPS 2026 · similarity 0.86)
- A Theory of Formalisms for Representing Knowledge (AAAI 2025 · similarity 0.86)
- Emergent Symbolic Mechanisms Support Abstract Reasoning in Large Language Models (ICML 2025 · similarity 0.85)
- When No Paths Lead to Rome: Benchmarking Systematic Neural Relational Reasoning (NeurIPS 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.
