Daniele-Cangi

Daniele-Cangi/Neural-Continuity

Evidence-driven framework for testing whether model transformations preserve declared operational behavior across conversion, quantization, and other transitions.

4 stars 0 forks 4 watchers Python MIT License

machine-learning ml-testing model-conversion model-validation onnx pytorch quantization reproducibility

5 Open Issues Need Help Last updated: Aug 16, 2026

Open Issues Need Help

View All on GitHub

Add an offline replay-only CI path for diagnostic components about 2 hours ago

enhancement help wanted good first issue

Daniele-Cangi/Neural-Continuity

4

Evidence-driven framework for testing whether model transformations preserve declared operational behavior across conversion, quantization, and other transitions.

Python

#machine-learning#ml-testing#model-conversion#model-validation#onnx#pytorch#quantization#reproducibility

Implement dominant-onset localization and frozen H1–H4 hypothesis reporting about 2 hours ago

enhancement help wanted

Daniele-Cangi/Neural-Continuity

4

Evidence-driven framework for testing whether model transformations preserve declared operational behavior across conversion, quantization, and other transitions.

Python

#machine-learning#ml-testing#model-conversion#model-validation#onnx#pytorch#quantization#reproducibility

Implement canonical diagnostic metrics and model-free replay about 2 hours ago

enhancement help wanted

Daniele-Cangi/Neural-Continuity

4

Evidence-driven framework for testing whether model transformations preserve declared operational behavior across conversion, quantization, and other transitions.

Python

#machine-learning#ml-testing#model-conversion#model-validation#onnx#pytorch#quantization#reproducibility

Capture paired FP32/INT8 activations under the frozen probe plan about 2 hours ago

enhancement help wanted

Daniele-Cangi/Neural-Continuity

4

Evidence-driven framework for testing whether model transformations preserve declared operational behavior across conversion, quantization, and other transitions.

Python

#machine-learning#ml-testing#model-conversion#model-validation#onnx#pytorch#quantization#reproducibility

Implement derived diagnostic graphs and exact instrumentation fidelity control about 2 hours ago

enhancement help wanted

Daniele-Cangi/Neural-Continuity

4

Evidence-driven framework for testing whether model transformations preserve declared operational behavior across conversion, quantization, and other transitions.

Python

#machine-learning#ml-testing#model-conversion#model-validation#onnx#pytorch#quantization#reproducibility