Research map / Probabilistic and scientific modeling / Molecules and proteins
Soft-Unification in Deep Probabilistic Logic
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 |
|---|---|---|---|
| replace probabilistic semantics with fuzzy Gödel t-norm (ablation) | DeepSoftLog · Countries S2 | AUC-PR: 97.67 ± 0.98 -> 23.73 ± 4.08 | Hurt |
| replace the hypersphere exp-angle soft-unification with a Gaussian kernel (ablation) | DeepSoftLog · Countries S2 | AUC-PR: 97.67 ± 0.98 -> 28.33 ± 3.64 | Hurt |
| use DeepSoftLog instead of NTP (probabilistic semantics, same templates) | NTP · Countries S3 | AUC-PR: 56.68 ± 17.6 -> 97.90 ± 1.00 | Helped |
| use DeepSoftLog instead of NTP | NTP · Countries S2 | AUC-PR: 87.40 ± 11.7 -> 97.67 ± 0.98 | Helped |
| use DeepSoftLog instead of Greedy NTP | GNTP · Countries S2 | AUC-PR: 90.82 ± 0.88 -> 97.67 ± 0.98 | Helped |
| use DeepSoftLog instead of Greedy NTP | GNTP · Countries S3 | AUC-PR: 87.70 ± 4.79 -> 97.90 ± 1.00 | Helped |
| use DeepSoftLog instead of Greedy NTP | GNTP · Countries S1 | AUC-PR: 99.98 ± 0.05 -> 100.0 ± 0.00 | No clear effect |
| use DeepSoftLog instead of A-NeSI | A-NeSI · MNIST-addition 15 digits | accuracy: 75.9 ± 2.2 -> 77.1 ± 1.6 | Helped |
| use DeepSoftLog instead of A-NeSI | A-NeSI · MNIST-addition 1 digit | accuracy: 97.7 ± 0.2 -> 98.4 ± 0.1 | Helped |
| use DeepSoftLog instead of Embed2Sym | Embed2Sym · MNIST-addition 15 digits | accuracy: 60.5 ± 20.4 -> 77.1 ± 1.6 | Helped |
| use DeepSoftLog instead of DeepStochLog | DeepStochLog · MNIST-addition 4 digits | accuracy: 92.7 ± 0.6 -> 93.5 ± 0.6 | Helped |
| increase digits to 100 with linear-scaling embedding encoding | DeepSoftLog · MNIST-addition 100 digits | accuracy: 25.6 ± 3.4 | Helped |
Closest papers by meaning
- Embeddings as Probabilistic Equivalence in Logic Programs (NeurIPS 2025 · similarity 0.89)
- DeepProbLog: Neural Probabilistic Logic Programming (NeurIPS 2018 · similarity 0.88)
- Improving Soft Unification with Knowledge Graph Embedding Methods (ICML 2025 · similarity 0.88)
- DeepStochLog: Neural Stochastic Logic Programming (AAAI 2022 · similarity 0.88)
- Learning Better Representations From Less Data For Propositional Satisfiability (NeurIPS 2024 · similarity 0.85)
- Right for the Right Reasons: Avoiding Reasoning Shortcuts via Prototypical Neurosymbolic AI (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.