63 configurations tested. Sweep results published with the model files.
Part of the RYS Sovereign Collection v1 — the headline crossover.
Where this sits in the Sovereign Collection
v1 — Qwen2.5 cross-scale + Qwen3-32B headline crossover. Four sizes from 0.5B to 32B; RYS works at every scale, with the lift size and dimension shifting by baseline:
v1 attribution: John Broadway, with collaboration from Claude (Opus 4.6 in April 2026 build; Opus 4.7 in May 2026 v1 republication aligning to the original Qwen2.5 cross-scale intent). Original RYS method by David Ng on Qwen2-72B; sweep toolkit by alainnothere.
v2 cross-architecture context (2026-05-13)
This model's place in the v2 curve: baseline reasoning 70.59%, peak RYS Δ +17.65%. The (20,28) 8-layer block sits at 30–50% depth, the family-invariant band for deeper models. 38 boosting configurations make this the most-validated single-model dataset in v2.
Pearson r(baseline reasoning, peak RYS lift) = −0.726. Weak baselines lift more, in their weakest dimension.
Three RYS-recoverable suppression mechanisms identified: under-training scale, MoE routing inefficiency, specialization training trade-off.
One published negative result (SmolLM2-1.7B). RYS is not universal.
v2 attribution: John Broadway, with cross-architecture analysis by Claude (Opus 4.7). Original RYS method by David Ng; circuit-finder toolkit by alainnothere.
