AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

Variational Learning for Insertion-based Generation

Yangtian Zhang, Zhe Wang, Arthur Gretton, Rex Ying, David van Dijk, Michalis Titsias, Jiaxin Shi

ICML 2026 · Probabilistic and scientific modeling › Molecules and proteins · approach: Neuro-symbolic probabilistic layers · 0 citations (OpenAlex)

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 changedWhereMeasuredEffect
learn the insertion policy instead of random insertion order (AO-IP) (ablation)IP · GuacaMol unconditionalvalidity (%): 88.2 -> 97.2Improved
learn the insertion policy instead of random insertion order (AO-IP) (ablation)IP · GuacaMol unconditionalvalid-unique-novel (%): 87.7 -> 94.9Improved
learn the insertion policy (ablation)IP · Fragment decorationvalidity (%): 40.9 -> 99.8Improved
learn the insertion policy (ablation)IP · Linker designvalidity (%): 50.6 -> 99.5Improved
learn the insertion policy (ablation)IP · Linker + full decorationunique valid (%): 31.1 -> 34.2Improved
use classifier-based instead of policy-based terminationIP · GuacaMol unconditionalvalid-unique-novel (%): 94.9 -> 95.6Improved
insertion-based IP instead of fixed left-to-right FO-ARMIP · Star graph planningsequence accuracy (%): 24.0 -> 83.0Improved
learned insertion order IP instead of LO-ARMIP · Star graph planningsequence accuracy (%): 30.1 -> 83.0Improved
IP instead of FO-ARM on distribution matchingIP · GuacaMol unconditionalvalidity (%): 98.5 -> 97.2Hurt
IP instead of FO-ARM on noveltyIP · GuacaMol unconditionalvalid-unique-novel (%): 80.4 -> 94.9Improved
relax insertion slot constraints (add decoration slots)IP · Conditional molecule generationunique valid (%): 13.1 -> 34.2Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy