Runs a fixed, int8-quantized logistic-regression-class credit-decisioning model as a pure integer inference kernel: an int8 weight vector, a fixed-point bias, and a threshold comparison, nothing else. Every input is a pre-normalized fixed-point integer and every operation inside the kernel is integer add, multiply, or compare, so the same score reproduces byte-for-byte on every compute surface.
🔒 All inputs are processed locally in your browser. No data is transmitted. Do not enter real personal data — use synthetic or anonymised inputs only.
⚠ This tool proves that THIS fixed quantized model produced THIS score from THESE inputs. It does not attest to the model's fairness, accuracy, or fitness for any real credit decision. The underlying model is a synthetic, offline demand-test artifact — not fit for regulatory credit decisioning. See the quantization_parity field on the exported artifact for the measured float-vs-quantized agreement rate.
Each value is the feature already normalized and scaled by 2^16 (a signed fixed-point integer), matching how the model was trained and quantized offline. Leave a field blank to treat it as 0.