Research map / Probabilistic and scientific modeling / Molecules and proteins
Learning a Generative Model for Validity in Complex Discrete Structures
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 mutual-information-based active sequence selection instead of uniform passive sampling | Python expression validity model · Python 3 expressions at temperature 0.100 | N: $9.7\times 10^{27}$ -> $8.2\times 10^{28}$ | Helped |
| Use mutual-information-based active sequence selection instead of uniform passive sampling | Python expression validity model · Python 3 expressions at temperature 0.025 | N: $2.9\times 10^{25}$ -> $4.3\times 10^{27}$ | Helped |
| Use mutual-information-based active sequence selection instead of uniform passive sampling | Python expression validity model · Python 3 expressions at temperature 0.005 | N: $1.1\times 10^{22}$ -> $1.3\times 10^{27}$ | Helped |
| Use mutual-information-based active sequence selection instead of uniform passive sampling | Python expression validity model · Python 3 expressions at temperature 0.100 | validity: 0.850 -> 0.841 | Hurt |
| Use mutual-information-based active sequence selection instead of uniform passive sampling | Python expression validity model · Python 3 expressions at temperature 0.025 | validity: 0.969 -> 0.995 | Helped |
| Use mutual-information-based active sequence selection instead of uniform passive sampling | Python expression validity model · Python 3 expressions at temperature 0.005 | validity: 1.000 -> 1.000 | No clear effect |
| Add a post-hoc validity model that masks invalid decoder choices | Plain CVAE · ZINC | sample validity: 0.5% -> 22.3% | Helped |
| Add a post-hoc validity model that masks invalid decoder choices | Plain CVAE · ZINC | reconstruction accuracy: 49.7% -> 50.2% | Helped |
| Use the post-hoc validity model rather than the Grammar VAE approach | Molecule VAE · ZINC | sample validity: 7.2% -> 22.3% | Helped |
| Use the post-hoc validity model rather than the Grammar VAE approach | Molecule VAE · ZINC | reconstruction accuracy: 53.7% -> 50.2% | Hurt |
Closest papers by meaning
- Steering Masked Discrete Diffusion Models via Discrete Denoising Posterior Prediction (ICLR 2025 · similarity 0.84)
- Bio-xLSTM: Generative modeling, representation and in-context learning of biological and chemical sequences (ICLR 2025 · similarity 0.84)
- Fine-Tuning Discrete Diffusion Models via Reward Optimization with Applications to DNA and Protein Design (ICLR 2025 · similarity 0.84)
- Beam Enumeration: Probabilistic Explainability For Sample Efficient Self-conditioned Molecular Design (ICLR 2024 · similarity 0.84)
- Universality of Linear Recurrences Followed by Non-linear Projections: Finite-Width Guarantees and Benefits of Complex Eigenvalues (ICML 2024 · similarity 0.84)
- Variational Search Distributions (ICLR 2025 · similarity 0.84)
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.