Research map / Probabilistic and scientific modeling / Variational and generative sampling
Discrete Object Generation with Reversible Inductive Construction
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 |
|---|---|---|---|
| apply reconstruction step to corrupted molecules (vs corrupted samples) (ablation) | GenRIC · ZINC molecules | QED: 21\% higher QED | Helped |
| run corrupter repeatedly with no reconstruction (ablation) | GenRIC · ZINC molecules | QED/SA distribution match | Hurt |
| increase mean corruption sequence length (ablation) | GenRIC · ZINC molecules | sample quality (QED/SA/logP distributions) | Hurt |
| shorten mean corruption length to one step (ablation) | GenRIC · ZINC molecules | SA distribution match | Helped |
| vary maximum molecule size from 25 to 35 atoms (ablation) | GenRIC · ZINC molecules | distribution statistics | No clear effect |
| vary training batch size to smaller values (ablation) | GenRIC · ZINC molecules | training result | No clear effect |
| use domain-specific validity-preserving moves vs GraphRNN | GenRIC · Laman graphs | % valid: 0.15 -> 100 | Helped |
| use GenRIC vs GraphRNN | GenRIC · Laman graphs | DoD KS distance: 0.96 -> 0.33 | Helped |
| GenRIC vs JT-VAE | GenRIC · ZINC molecules | QED KS distance: 0.090 -> 0.045 | Helped |
| GenRIC vs JT-VAE | GenRIC · ZINC molecules | SA KS distance: 0.21 -> 0.28 | Hurt |
| GenRIC vs SMILES LSTM | GenRIC · ZINC molecules | SA KS distance: 0.051 -> 0.28 | Hurt |
| GenRIC vs CG-VAE | GenRIC · ZINC molecules | SA KS distance: 0.56 -> 0.28 | Helped |
Closest papers by meaning
- Discrete Markov Probabilistic Models: An Improved Discrete Score-Based Framework with sharp convergence bounds under minimal assumptions (ICML 2025 · similarity 0.87)
- Flow Matching with General Discrete Paths: A Kinetic-Optimal Perspective (ICLR 2025 · similarity 0.86)
- Generative Flows on Discrete State-Spaces: Enabling Multimodal Flows with Applications to Protein Co-Design (ICML 2024 · similarity 0.86)
- Spherical Flows for Sampling Categorical Data (NeurIPS 2026 · similarity 0.85)
- Unlocking Guidance for Discrete State-Space Diffusion and Flow Models (ICLR 2025 · similarity 0.85)
- Steering Masked Discrete Diffusion Models via Discrete Denoising Posterior Prediction (ICLR 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.