README.md

metadata

title: SPARK
emoji: ⚡
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: 5.29.0
app_file: app.py
pinned: false
license: mit
short_description: SPARK — Bayesian inference for CV & TPD analysis

SPARK has moved

This Space has moved to https://huggingface.co/spaces/bing-yan/spark. The model, code, and inference behaviour are identical — only the URL and branding (TRACE -> SPARK) have changed. This /trace URL remains live as a redirect for existing bookmarks.

SPARK — Simulation-based Posterior Amortization for Reaction Kinetics

Amortized Bayesian inference for electrochemistry and catalysis. Upload cyclic voltammetry (CV) or temperature-programmed desorption (TPD) data to automatically:

1. Identify the reaction mechanism from a candidate list (9 for CV, 11 for TPD)
2. Infer kinetic parameters with full Bayesian posterior uncertainty

Inference takes ~50 ms on CPU. Accepts CSV files or plot images.

The deployed checkpoints are the noise-augmented headline models (CV: v14_9mech, TPD: tpd_11mech_v2) that retain 92.4 % (CV) and 95.6 % (TPD) classification accuracy on realistic noisy inputs.

Supported Mechanisms

Electrochemistry (CV, 9 mechanisms): Nernst, Butler–Volmer, Marcus–Hush–Chidsey, Adsorption, EC, Langmuir–Hinshelwood, EE, EC′, CE.

Catalysis (TPD, 11 mechanisms): First-order, Second-order, Zeroth-order, FirstOrderCovDep, DiffLimited, Precursor, Dissociative, ActivatedAdsorption, LH Surface, Mars–van Krevelen, TwoSite.

Citation

[Citation to be added upon publication]