research
One Neural Pipeline Chooses the Reaction Path and Its Rates
Shared sequence encoders support both model selection and parameter inference for competing chemical networks.
Summary
Shared sequence encoders support both model selection and parameter inference for competing chemical networks.
The architecture combines one-dimensional convolution, set attention, a model-selection network and conditional invertible networks. A protodeboronation example shows how it handles ambiguous catalyst-deactivation pathways. The paper demonstrates inference on the studied system rather than general chemical identification.
Why it matters
Shared sequence encoders support both model selection and parameter inference for competing chemical networks.
Limits and context
No additional limitation was separately recorded.
Key claims
Shared sequence encoders support both model selection and parameter inference for competing chemical networks.
Evidence: source-2026-10-14-020
Sources
- arXiv preprint 2610.18825, original v1arXiv / submitting research authors · primary research
Corrections
No corrections have been recorded for this story.