Ignis Cogitationis

This repository contains the model checkpoints, downstream evaluation scores, and pretraining convergence logs for the Quantum Like Attention Framework (Q.L.A.F) 1.3B configuration. Pretraining is executed under the Odyssey Route-B Unified Engine with PyTorch Distributed Data Parallel (DDP) scaling support. $$\text{entropy\weight} = 0.03 \times \left(1.0 - \frac{\text{step}}{20000}\right)$$ This forces the router to explore uniform row distributions, preventing it from locking onto random rows early and guaranteeing convergence to the correct state updates for secret key retrieval. Evaluations are conducted across three independent seeds (42, 100, 2026). Accuracies are reported on MMLU (500…