cl-metrics: A Stateless Python Library for Continual Learning Evaluation with SNN Energy-Aware Extensions

cl-metrics is a stateless Python library that computes all standard Continual Learning evaluation metrics — Average Accuracy (AA), Backward Transfer (BWT), Forward Transfer (FWT), and others — from a raw N×N per-task accuracy matrix, with no dependency on any training framework, eliminating the silent formula drift that makes results across papers subtly incomparable. The library also introduces the first standardised evaluation suite for Spiking Neural Network continual learning, comprising four energy-aware metrics grounded in spike firing rates that quantify the accuracy-energy tradeoff that accuracy alone cannot capture. All implementations follow their original published formulations exactly, and every result is validated against the Maya Research Series (Swaminathan, 2026a–2026g), a seven-paper neuromorphic SNN continual learning benchmark on Split-CIFAR-10 and Split-CIFAR-100. This work is part of the Maya Core series — thirteen papers implementing the Advaita Vedantic Antahkarana as computational primitives in spiking neural networks — with the Bhaya Quiescence Law (β* ≤ 0.32%) and Buddhi S-Curve Determinism (R² = 1.0000) confirmed across all papers in the series. Series: Part of the Maya Research Series — 13 papers implementing the Advaita Vedantic Antahkarana as computational primitives in spiking neural networks. Bhaya Quiescence Law (β* ≤ 0.32%) and Buddhi S-Curve Determinism (R²=1.0000) confirmed across all papers. Links: GitHub Repository (private — to request access: email research@nexuslearninglabs.in with subject Code Access Request — cl-metrics and your research context) | FAQ | Full Series Index — venky2099.github.io Nexus Learning Labs, Bengaluru · UDYAM-KR-02-0122422 · BHASKAR IN-0526-9452JSORCID: 0000-0002-3315-7907 · VAIRAGYA_DECAY_RATE = 0.002315 (embedded in all canonical hyperparameters)Canary: MayaNexusVS2026NLL_Bengaluru_Narasimha