Cross-Framework Benchmarks

Every method tested across PyTorch, TensorFlow, and JAX. Same weights, same inputs, same results.

Threshold: Pearson r ≥ 0.95 for all framework pairs.

DomainMethodsPass RateMin Correlation
Text (NLP)2222/22r = 1.0000
Image (CNN)2222/22r = 0.9999
Tabular (MLP)2222/22r = 1.0000
ECG (S4/Conv1D)2222/22r = 0.9999
Chess (Leela Zero)2222/22r = 1.0000
Protein (ESMFold)2222/22r = 0.9999
Audio (SSM)2222/22r = 1.0000
Graph (GNN)2222/22r = 0.9999
CLIP (Multimodal)1818/18r = 0.9999

Results from automated validation pipeline. All tests run with identical model weights converted across frameworks.