route2's plotted hack/solve is the DEPLOY eval (hk_dep/slv_dep): greedy, n=64,
logged every eval_ablate_every=5 steps and EMA-held flat between. The held line
reads as per-step-dense and oversells route2's smoothness vs the per-step
temperature-sampled (n=28) training curves the other arms plot -- an apples-to-
oranges smoothness the reader shouldn't be misled by.
_mark_if_sparse dots the real measured points when a series is >50% NaN; dense
series (training hack_s, cos sep/leak) stay unmarked. Now the route2 curve
visibly rests on ~13 eval points, not 60.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The cos row mixed non-comparable quantities: erase logged cin_t/cin_s
(pre-intervention single cosines), route2 logged hkgap (a difference) and
resid (post-intervention), all under one "cos(grad,v_hack)" ylabel. Wassname
flagged it -- they are not the same measurement.
Derive two quantities that mean the same thing in every column (_add_cos_derived):
sep = does v_hack still discriminate hacky from clean gradient
erase: cin_t-cin_s (teacher vs student); route2: hkgap (hackflag vs clean)
leak = residual hack-alignment of the post-intervention deployed gradient
erase: cout (after projection); route2: resid (after routing) -- same quantity
Legend now lands on the leftmost arm that has cos data (vanilla has none).
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Stacked hack (top) + solve (bottom) sharing x; EMA-5; onset dot on hack only;
arms direct-labelled once on solve with y de-collision + leader lines (the three
non-route arms overlap, so their labels would otherwise stack). routing2 reads
hack~0 / solve highest at a glance.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
- plot_dynamics: routing (route v1) out of ARM_ORDER -- superseded by routing2.
- plot_substrate: per-mode hk_* are now plain per-batch counts (streaming log
dropped the /denominator); parse the count, plot it (EMA or cumsum); skip old
n/d-format logs (incompatible units). Y-axis hacks/batch, count annotations.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The plotter picked hk_abl (dense proxy) whenever the COLUMN existed, but no-floor
runs (rollout_ablate_frac=0) emit hk_abl as 0/0 -> all-nan, so the deploy panel
came up empty. Test for finite data (_has_data) not column presence; fall back to
the sparse-but-real hk_dep (every eval_ablate_every steps). _ema carries values
across the nan gaps -> a held step-line.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Behavior-safe: comments/docstrings only. smoke + smoke-route2 exit 0, metrics
identical. Clears the 26 comment em-dashes in proj/rewards/extract_vhack_grad/
probe_distill/regrade_pool/verify_vhack_heldout/probe_plot_stack/pairs.
One em-dash deliberately preserved: pairs.py:313, inside a contrastive-pair
completion string ("# Sample inputs — uncomment ..."). It is training data
(feeds v_hack extraction), not code style, so `grep -P '—' src/` bottoms out
at 1 rather than 0. Changing it would alter the experiment's inputs.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Behavior-preserving (smoke + smoke-route2 exit 0, metrics identical, route2
‖δS_hack‖=0.0079>0). All touched modules import-checked (no cycles).
Mirrors the clean repo's responsibility split:
- ref_logprobs_via_zero_delta + ablate_quarantine -> antipasto.py (the adapter
owns the δS=0 free-ref-model trick and the δS_hack ablation).
- load_v_hack + postprocess_v_hack -> extract_vhack_grad.py (alongside extract_v_hack).
- load_problems + DATA + the per-mode hints -> new problems.py.
Importers updated to the new homes (probe_distill, derisk_loopholes,
verify_vhack_heldout, probe_lora_runtime, build_substrate, regrade_pool,
scripts/validate_spoonfeed). Moving DATA out of train.py also broke the
regrade_pool->train edge, so train.py can now import the v_hack helpers at
top level without a cycle.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
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>
The cumulative n/d (e.g. 14/52) grew unboundedly and read as noise. Now
each hk_<mode> shows just this step's student hacks of that mode as a plain
int. The running mode_hacks/mode_rollouts tallies still feed the end-of-run
substrate learning table.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
--latest-per-arm + --min-steps select the freshest >=N-step log for each
arm from logs/, no hand-globbing. Harden parse_log against historical logs:
require '| INFO |' in the header line, drop pure-symbol header tokens.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The routing arms' benefit shows on the DEPLOYED model (quarantine deleted).
Prefer the dense per-step proxy hk_abl/slv_abl (every step, rollout_ablate_frac>0)
over the sparse held-out hk_dep eval for the plotted hack_s/gt_s curve; fall back
to hk_dep for runs that predate the proxy.
- parse hk_abl/slv_abl; routing+routing2 substitute it (else hk_dep) into hack_s/gt_s
- classify/ARM_ORDER/ARM_COLORS recognise routing2
- gate cos cols (cin_t/cin_s) by presence: vanilla/routing2 lack them, so parse
and panels skip them instead of KeyError (also fixes a pre-existing vanilla crash)
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The ablated rollout slice IS deployed-model samples (quarantine off), so grade
it for a free per-step deploy hack/solve proxy -- no extra generation, unlike
the periodic eval_ablate_every eval. Distinct columns from hk_dep/slv_dep
(held-out greedy = plot number); the proxy is on train prompts at sampling
temp, noisier same-distribution.
- rollout_ablate_frac default 0.0 -> 0.5 (the exploration floor, now on by default)
- gen_students returns (rows, n_abl); mark the ablated tail; agg_is_ablated
- hk_abl/slv_abl columns (routing/routing2 arms)
- SmokeConfig group 2->4 so route2 smoke (mix=0.5 -> G_s>=2) can split the slice
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Generate a fraction of student rollouts with delta_S_hack ablated (deployed
model -> can't hack -> explores solves), so the solve region stays covered
even if on-policy sampling collapses onto hacking. Motivated by job 60's
hkgap decay to ~0 post-emergence (gate stops discriminating; risk that hack
eats everything and delta_S starves). Pure sampling-side diversity, no
no-cheat-boundary impact; frac=0 = unchanged. Smoked at frac=0.5.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Disambiguates qE: high qE + resid~0 = hack stripped cleanly into the
deleted quarantine; high qE + resid>0 = false-negative leak into the
deployed knob. hkgap measures cloud separation, qE measures energy,
neither measures hack-ward alignment of the KEPT gradient. (DeepSeek
review #5 starvation-vs-correct-routing diagnostic.)
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The CPU smoke ran fp32 + sdpa, so it never walked the bf16/flash_attn2 path the
real run uses -- a whole dtype/magnitude bug class was invisible to the gate (per
the smoke principle: a path that doesn't fire in smoke isn't covered). The tiny-
random model peaks ~1.4GB on GPU, so cost is negligible. Drop CUDA_VISIBLE_DEVICES=
from every smoke recipe; train.py auto-detects cuda -> bf16. (Stale fp32 smoke
v_hack must be re-extracted bf16; auto-extracts on cache-miss.)
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The distinct-basis A_q/B_q LoRA (~33M params at rank-16) gave the quarantine a
~100x capacity edge over delta_S, so routing-everything-there was the low-
resistance path: qE pinned ~0.97 (energy into the thrown-away knob) while the
deployed delta_S learned nothing (job 54). The cause was capacity imbalance, not
the routing gate (calibrated-tau already separated hack/clean, hkgap>0).
Consolidate to one adapter type: the quarantine is now delta_S_hack, the second
diagonal in the same frozen SVD basis, shape [r], capacity-matched to delta_S,
zeroed at deploy. route2's calibrated-tau gate parks the flagged rollouts' grad
into delta_S_hack.grad (like proj.py's route parks its subspace projection);
delta_S keeps the unflagged. Both diagonals train at one shared lr.
Removed: A_q/B_q params, v_act + extract_v_act, the act-mask arm (a shared
diagonal can't be per-token gated), route2_mask / route2_quarantine_rank /
route2_quar_lr_scale knobs, the separate quar optimizer group. Arm name
routing2_{act,grad} -> routing2. v_grad refresh extracts from delta_S (main)
with the quarantine ablated.
SGTM check: their gradient routing uses a hard detach on capacity-matched
reserved dims, no soft/tanh/sigmoid gate -- balance is the fix, not gating.
Smoked clean: tau/hkgap/qE render, ||delta_S_hack||>0 assert passes, exit 0.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
tau = EMA midpoint of hack-cloud (teacher + detector-flagged student) vs
clean-cloud (not-flagged student) cos(g_b,v_grad), per module. Rides the cin
drift; force-routes known hacks, tau-routes the ambiguous rest (incl unknown B).
New cols tau + hkgap (hack-clean separation gauge). Keeps the vector premise --
the flag only calibrates, never gates. Spec: docs/spec/20260601_calibrated_tau_route2grad.md
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Routing stays vector-based (cos>tau, not the detector flag) but tau is the
per-step EMA midpoint of the hack vs clean cos clouds (teacher+flagged-student
anchor hack; not-flagged anchor clean). Rides the cin drift; force-routes known
hacks; tau-routes unknown B. Logs tau + hkgap. No-cheat: detector only
calibrates, gt_pass never gates.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
DeepSeek/GPT-5.5/Gemini converge: (1) UNANIMOUS top concern -- prove the v_hack
DIRECTION is causal, not the detector flag/capacity (random-V + flag-only triad);
(2) route2-grad over-routes too (cos>0 = ~50% coin-flip by concentration, not a
granularity fix); (3) improvement B != erase only via on-policy generation, which
ablate-during-gen would remove.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The audited last-gen alone has no reference. A frozen coherent vanilla snippet
(maxPoints step 59) above it makes salad obvious -- e.g. job 46 step 14 is
clearly soup next to it, even though lp_t stayed flat and the tripwire missed it.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The per-refresh logger.info was noise (one line per refresh @every-2-5 steps).
The refr column already marks the refresh; for act-mask it now carries the
basis-overlap (mean |cos| old-vs-new v_act). grad-mask keeps a bare marker.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
- route2 v_act/v_grad refresh now logs basis_overlap_with_prev (mean |cos| of
old vs new mask direction) -- matches the clean-repo guard; a bare refresh bool
carried no info, overlap shows if the mask chases a drifting target.
- divergence tripwire gets a soft logger.warning at 3-nat lp_t drop before the
5-nat hard abort (early 'coherence slipping, lr too high?' heads-up).
- threshold note: healthy lp_t runs -0.5..-2.5, collapse ~-11, so an absolute
<-1 warning would false-fire; relative-drop-from-best is the right test.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
route2-act diverged (run 43): 33M kaiming A_q/B_q at delta_S's lr=3e-3 blew up
(gn 0.3->7.5 step 8, generations -> token salad, lp_t -11). Fixes:
- #167 separate quarantine lr (route2_quar_lr_scale=0.1) so the 60x-bigger fresh
LoRA isn't trained at the main-knob lr.
- #168 divergence tripwire on teacher ppl (lp_t high-water mark; abort if it
drops >5 nats for 2 steps). Relative so tiny-random smoke (flat lp_t~-11.9)
doesn't false-trip.
- #165 act-path was silent: stash cos(a,v_act) + fired-fraction in the forward,
surface as act_cos/act_fire columns (route2-act). smoke shows act_fire=0.64 =>
the cos>0 sign test over-routes (fires on most tokens, not just hack ones).
- #166 print last train generation before FINAL EVAL (coherence eyeball).
- route2 v_act/v_grad refresh was firing but silent -- now announced.
- #162 plot_deploy_overlay.py: per-mode DEPLOY overlay from per_mode_deploy.json
(honest shipped-model numbers, route2-safe). just plot-deploy.
- just plot/results hardened: parse by header name, skip non-substrate logs,
non-fatal aggregate delegation.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Old GT_S=6/HACK_S=8 were the pre-sprd/N layout; current table is gt_s=4
hack_s=6, so newer logs were silently mis-read and old distill logs crashed
_frac on a non-fraction token. Now locate the train.py streaming header
(first token 'step' + 'ref_eq' present) and map columns by name.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Three route runs all show deleting the quarantine raises solve and lowers hack.
Mechanism: clean-rollout solve gradient stays unflagged -> flows to delta_S; the
hack masks that competence at train time, revealed at deploy. Exception: run_tests
(solve 0->0) where hacking fully dominated exploration. Logs the 3 failure-mode
checks (eval artifact / teacher-distillation / random-V null).
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
#164: the final eval now runs for ALL arms (not just route/route2) on the
same fixed eval subset, so the all-arms overlay reads identical per-mode
numbers. vanilla/erase have no quarantine -> deploy == train (one eval);
route/route2 also run the knob-off (ablated) eval. Writes a single
per_mode_deploy.json into run_dir (arm, mask, refresh, seed + per-mode
train/deploy hack+solve) as the canonical source for the #162 overlay plot.
justfile: replace the parametrized run-substrate (which re-passed seed/steps/
refresh/mask defaults every invocation) with one explicit queue-substrate that
queues the fixed 5-arm overlay set, each arm passing ONLY its non-default flags.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The verbose argv (--teacher-pool-dir, --vhack-pairs-path, and redundant
--vhack-refresh-every/--seed/--steps) came from run-substrate passing
everything explicitly. steps/seed/refresh were already defaults; the two
paths weren't. Now FastConfig defaults to the current experiment line so a
real run needs only --intervention (+ optional seed/refresh/mask). Smoke
(SmokeConfig) unaffected -- it sets its own pool. Stripped the recipe to match.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Smoke is fp32 (CPU tiny-random) so the bf16 path never fired -- job 34/35
crashed on the real Qwen3-4B with 'BFloat16 != float' in the quar matmul.
Cast A_q/B_q/v_act down to activation dtype in the forward, mirroring the
delta_S.to(a.dtype) pattern (fp32 master, bf16 compute, grads cast back).
Validated forward+backward in bf16 for both masks. + run-substrate MASK param.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
route2 uses v_act/v_grad, not v_hack, so --vhack-refresh-every never fired
for it -- the mask was frozen regardless of the flag. Frozen real-V route
(job 32) shows why this matters: cin_t decays to cin_s by step 7, deploy hack
only drops ~8pp (vs run-31 rf5 ~0). Now re-extracts v_act/v_grad every N steps
with the quarantine ablated (same MASK_PAIRS, no oracle). + journal entry (j).
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The hook gate is necessarily per-token ([G*s, r], nn.Linear flattens the
batch). _route2_grad_filter now sums each rollout's token gate-grads before
the cos(g_b, v_grad) flag, so routing is per-rollout (the preregistered GRPO
unit) and the sign is denoised. Per-token a clean rollout scatters ~50% of
tokens over cos>0 by noise, spuriously routing half its gradient mass.
Verified by deepseek-v4-pro review: gate identity, divide-out, eps-guard,
Arm B detach-route, R5 no-cheat all correct; this was the one finding.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Arm A (route2_mask=grad): per-rollout gate splice (identity at c=1) recovers
the per-sample delta_S grad after backward (c.grad = delta_S * g_b); train.py
divides it out (eps-guard |delta_S|>1e-6), flags rollouts by cos(g_b, v_grad)>0,
and SUBTRACTS them from delta_S.grad. Single-pass, no forward detach, no second
backward -- the cross-step mismatch that made the spec's A1 stale-mask awkward
never arises (routing is post-backward within the step). v_grad = unit-mean
gradient diff from extract_v_hack raw grads (gradient-space analogue of v_act).
route2 forces the combined (non-split) backward since cos_pre is NaN for it
anyway, which also gives the gate a single clean grad to read.
Drop route2_tau: never tuned; the mask is cos>0 (the natural hack-ward boundary)
and the load-time noise floor already filters axes.
v_hack path now auto-derives from --vhack-pairs-path (out/vhack/v_hack_pairset_
<stem>.safetensors): pass the pairset, the hack file auto-loads/extracts -- no
need to also pass --v-hack-path. run-substrate drops the redundant flag.
smoke: smoke-route2 (act) and new smoke-route2-grad both pass (||B_q||=0.109,
exit 0); erase shared-basis path unchanged (cout->0, fired~0.9).
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Routing-v2 spec (distinct-basis quarantine, two arms, proofs); related-work
no-cheat scorecard for TDGA/Cloud/SGTM/Confessions; full-text fetches of the
Deng and SGTM papers; journal entry for the run-31 confound + T1/T2 landing.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
Adds intervention=route2: a LoRA quarantine (A_q,B_q) with its own basis,
always summed into the forward, plus a per-sample activation-cosine mask that
detaches the kept adapter for flagged samples. Routing happens in the forward,
not via grad surgery: a flagged sample updates only the quarantine; an unflagged
hack-like sample concentrates there by gradient magnitude (absorption). Deploy
zeroes A_q,B_q. v_act built by extract_v_act (forward-only activation mean-diff
over persona pairs). Fixes the per-prompt zero_grad wiping quarantine grads
before opt.step. scripts/make_random_vhack.py = the random-V route control.
vhack_refresh_every default 0->5 (0 is ablation-only).
Smoke: R1 grad check passes (flagged->delta_S grad 0, A_q/B_q>0; forward value
unchanged); smoke-route2 ||B_q||=0.109, deploy eval + asserts pass.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
The signed sum(c)/||g|| let +/- v_hack axes cancel, reading ~0 even while a
large hack-ward magnitude was being routed -- a misleading gauge that drove
the 'route does nothing' misread. relu(c) BEFORE the norm matches what the
one_sided projection actually removes (||removed||=||relu(c)|| for orthonormal
V), so cin reads as 'fraction of grad stripped' in [0,1] and cout -> 0 exactly
after erase. Renamed _signed_cos -> _hackward_cos; flagged the now-invalid
E[cos|clean]=0 decomposition in probe_plot_stack.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
- plot_substrate.main now also calls plot_dynamics.plot/plot_hack_overlay so one
command produces all 4 figs (by_method, by_hack, aggregate, hack_overlay); the
aggregate 'total hacks per arm' core plot is kept, not reimplemented.
- plot_dynamics: point parser at CURRENT streaming headers (cin_t/cin_s, hk_dep/
slv_dep); it was built for the old cos_pre_t/hack_deploy spelling and silently
failed on sub4 logs. No backward-compat for the superseded header.
- justfile: 'plot GLOB STEM' canonical entrypoint over logs/*_sub4_*.log.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
eval_hack_solve now returns by_mode tallies; the final ROUTE EVAL logs
deploy hack/solve per env_mode, tagged IN-dist (run_tests, what v_hack saw)
vs held-out (file_marker/sentinel/stdout_marker). Distinguishes 'route
suppressed only its own mode' from 'route generalised to unseen modes'.
Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
