CLI toolkit that ingests qk-sniffer dumps, measures per-head positional predictability and attention plasticity, and exports CSV stats plus ready-to-share plots.

Automates attribution-graph analysis via probe prompting: circuit-trace a prompt, auto-generate concept probes, profile feature activations, cluster supernodes.

This repository introduces an adaptive formula inspired by the CHSH logic, designed to evaluate, test, and improve model performance across multiple conditions. By adapting CHSH principles into a flexible structure, it provides a systematic way to analyze results, ensure reliability, and explore deeper insights in experimentation.

Chiral Narrative Synthesis workspace for Thinker/Tinker LoRA pipelines, semantic fact-checking, telemetry, and reviewer-ready CNS docs.

Digital twin & security validation toolkit for DV-QKD (BB84 + decoy)

Lean 4 library + CLI for rigorous bounds in transformer computations (mechanistic interpretability)