Mehmet Arda
Hakbilen

My main focus is trainscope: a post-mortem flight recorder for LLM training loss spikes. Its CUSUM change-point detector catches gradual loss drift 5-20 steps before it turns into a full spike, validated with 0.0% false positives across 280+ noise scenarios and 100% sensitivity for 0.10σ-0.50σ drift. When failure occurs, it reconstructs which layer moved first, whether from activation kurtosis or gradient explosion.

Trainscope grew out of work on ENGRAM, a hybrid sequence backbone interleaving Mamba-2 SSD blocks with Gated Delta Rule blocks (3:1 ratio). It applies identical hyperparameters to 12-lead ECG, spoken commands, and sequential images, verified end-to-end across 312 numerical equivalence and streaming tests.

Earlier, I closed out NOESIS, testing if Titans-style associative memory matrices could inject continual learning directly into a frozen LLM manifold. The result was negative: frozen manifolds cannot assimilate untrained vector injections without distortion (0% recall on a 50-fact benchmark; full analysis in FINDINGS.md).