default rollout_ablate_frac=0 (drop the exploration floor)

job 64 (frac=0.5) leaked: deploy hack climbed 0.00->0.125 over training while
no-floor job 60 held 0.000. Sampling from the deployed model makes its reward an
optimization target whose optimum is hacking, so stale-gate misses update the
kept knob -- the floor optimizes the policy the quarantine should keep clean.
Neither gradient-routing paper samples from the ablated model in training. Off by
default; re-enable only to test floor+fresh-gate (refresh-every-1).

Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>

src/projected_grpo/train.py

@@ -175,18 +175,20 @@ class Config:
     project_overshoot: float = 1.0    # remove overshoot*c_use@V; 1.0=just remove, 1.1=10% reversal of hack-ward grad
     # Exploration floor against hack-saturation (route/route2 only). Fraction of
     # student rollouts to generate with the quarantine (delta_S_hack) ablated, i.e.
-    # from the DEPLOYED model. The risk this guards: if on-policy sampling collapses
-    # onto hacking, the policy stops emitting solves, every rollout gets routed to
-    # the quarantine, and the deployed delta_S never sees a solve gradient to learn
-    # from (it saturates). Forcing a fraction of rollouts hack-OFF guarantees the
-    # solve region stays covered, exactly like any RL exploration term. Pure
-    # sampling-side diversity; accepts a slight off-policy mismatch (GRPO already
-    # tolerates it) in exchange for guaranteed coverage. 0 = off.
-    # Default 0.5 (half the rollouts deploy-mode): the bonus is that these ablated
-    # rollouts ARE deployed-model samples, so we grade them for a FREE per-step
-    # deploy proxy (hk_abl/slv_abl) -- see the row dict -- without the extra
-    # generation the periodic eval_ablate_every eval costs.
-    rollout_ablate_frac: float = 0.5
+    # from the DEPLOYED model. Intent: if on-policy sampling collapses onto hacking,
+    # every rollout gets routed to the quarantine and the deployed delta_S never sees
+    # a solve gradient (it saturates); forcing some rollouts hack-OFF keeps the solve
+    # region covered.
+    # DEFAULT 0 (off). job 64 (frac=0.5) leaked: deploy hack climbed 0.00->0.125 over
+    # training while the no-floor job 60 held deploy hack=0.000. Mechanism: sampling
+    # from the DEPLOYED model makes its reward an optimization target whose optimum is
+    # hacking, so whenever the (stale) routing gate misses, those deploy-generated
+    # hacks update the kept knob -- i.e. the floor optimizes the very policy the
+    # quarantine is meant to keep clean. Neither gradient-routing paper samples from
+    # the ablated model during training (they sample knob-on, ablate at eval, and let
+    # absorption localize). So default off; re-enable only to TEST the floor+fresh-gate
+    # hypothesis (does refresh-every-1 keep the gate sharp enough that the floor is safe?).
+    rollout_ablate_frac: float = 0.0
     # Which grader flaw + factual hint this run trains on (a "hack class"). Sets
     # the prompt hint (HINT_REPLACE_TO) and how `passed` is graded in rewards.py.
     # run_tests = the original run_tests-overwrite loophole. eq_override / exit_code