Files
evil_MoE/docs/spec
wassname d6342ab201 feat: gradient routing — delta_S_hack quarantine + intervention {none,erase,route}
Stage-1 (T3) of the routing spec. Adds a per-module quarantine knob
delta_S_hack (AntiPaSTO forward = delta_S + delta_S_hack, both 0 at init).
intervention=route parks the hack-ward grad component (g - cV to delta_S,
cV to delta_S_hack) instead of erasing it; eval ablates delta_S_hack.

- proj.py: route flag splits the grad (overshoot=1, no rescale -> the split
  sums to g, so the training forward still moves hack-ward; route ⊇ erase).
- antipasto.py: second trainable knob, identity preserved at init.
- train.py: arm -> intervention {none,erase,route}; arm kept as a derived
  display name so run-id/BLUF/results.py/plot classify are unchanged. opt
  steps both knobs (hack knob grad=None under none/erase -> AdamW skips it,
  so erase reproduces old `projected` bit-for-bit, R4). R3 span assert
  (resid/||gh|| < 1e-4) + end-of-run ||delta_S_hack|| guard (route >0).
- results.py / plot_dynamics.py: read arm from the preset line (covers both
  old --arm and new --intervention logs); plot classifies `routing`.

smoke: none ||dsh||=0, erase clean, route ||dsh||=0.0105 span=2.9e-7. 64
archived projected rows still parse.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-05-30 00:31:30 +00:00
..
wip
2026-05-28 12:44:20 +00:00
2026-05-29 05:42:28 +00:00
2026-05-29 06:29:43 +00:00