Old filter used to_do_has_positive_honesty / not_to_do_has_negative_honesty
flags, which kept 194 to_do + 33 not_to_do = 227 rows (85% to_do only).
The dataset has honesty_label=+/-1 for all 438 rows via symmetric labeling
(if to_do is honest, not_to_do gets -1 regardless of its own values).
Filtering to honesty_label != 0 keeps all 438 rows and pairs both sides of
every dilemma, matching AntiPaSTO's symmetric evaluation approach.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Was: "Do you choose to: {action}?" with "My choice: **" anchor.
AntiPaSTO ref: "Should you take this action? Action: {action}" with "My choice:" anchor.
The ** anchor told the model to produce markdown-bolded answers, scoring
at the wrong token boundary. "My choice:" anchor scores " Yes"/" No"
which get_choice_ids already catches via prefix match.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Gemma-3/4 don't have </think> as a special token, so guided_cot_one
raised RuntimeError and killed the whole sweep. Fix: add has_thinking_mode
to _tok_extras and gate phase_a2 in replicate.py on it.
README cut from ~380 to ~120 lines: results tables, how to run, cite, links.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Gemma 4 (E2B/E4B) uses channel-based thinking tokens (<|think|>, <|channel>).
chat_template_extras() detects this via template string and passes
enable_thinking=False to all apply_chat_template calls in data gen,
dilemmas eval, and KL calib (via build_chat_text). Qwen3 and Gemma 3
return {} (existing thinking-mode handling unchanged).
Refactor calibration to match the gist methodology: for each prompt, greedy-
generate n_tokens under the steered policy, capture per-step steered
log-probs, then teacher-force the same continuation under base. Per-position
KL(steered ‖ base) is computed along the steered trajectory rather than at
fixed continuation positions. Captures cumulative drift the old fixed-
continuation KL missed.
Replaces 1-step Newton on alpha with exponential bracket -> Illinois regula-
falsi in log-(alpha, p95) space. Linear in log-log (since p95 ~ alpha^k near
root) so usually converges in 3-4 iters, with Illinois rule breaking the
stuck-endpoint failure mode of pure regula falsi.
Adds SI_to_do, SI_not_to_do (and si_fwd/si_rev/n_cho/n_rej variants)
to compute_full_metrics output. Lets us see if a steering method's
effect is symmetric across "Should you do X?" vs "Should you NOT do X?"
framings, or if it leans on one. Already-known pattern: paired-opposite
filler rows produced large SI_to_do > SI_not_to_do gaps before the
honesty-only filter landed.
The wassname/daily_dilemmas-self-honesty dataset uses paired-opposite
labels: if to_do has honesty in you_values -> to_do=+1, paired not_to_do=-1
even when not_to_do's own values are e.g. ['empathy'], unrelated to
honesty. Those 211/438 filler rows were dragging dW SI negative.
Changes:
- dilemmas.py: filter to action-specific honesty rows at load (227/438)
- data.py: narrow training pair to ('an honest',) / ('a dishonest',),
matching RepE extraction, simple_*_prompt baselines, and AntiPaSTO
honesty synonyms. Cluster paraphrases (candid/sincere/manipulative/...)
drift onto a broader axis -- comparison was unfair.
- prompt_baseline.py: regenerate engineered prompts via AxBench J.2
template with narrow-honesty concept slot. Old prompts brought in
integrity/transparency (POS) and Machiavellian/manipulation (NEG).
- nbs/rescore_honesty_only.py: ad-hoc rescore script confirming the
fix flips dW from -16 SI to +60 SI without rerunning the model.
α=1 means very different things across LoRA/PiSSA/DeLoRA/OFT/IA3/RepE/prompt;
calibrate α per method so p95 token-KL on held-out continuations matches
prompt:engineered_prompt_honest's footprint (≈0.61 nats over 50 stratified
prompts, 100 audit). Newton iter α_next=α·sqrt(T/M) converges 7/7 methods
in 2-3 iters. At calibrated ±α on daily-dilemmas (n=219), all 6 adapters
land deeply negative SI: fix counts cluster at 14-19 across all methods,
but adapters break 65-139 already-honest rows (vs 15-20 for engineered
prompts). Interpretation: prompts perturb topic-conditionally, adapters
uniformly — at matched off-task budget, adapters scatter mass over
already-correct rows. RepE sits between.
Caveats: single seed, calibration off-task, anchor audit p95 is 1.78×
calib (calibrated conservatively).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Address pi-review issues:
- SI_best: max(si_fwd, si_rev) does not equal "best honesty under post-hoc
sign flip" because under k_fpr=2 the FPR penalty hits the swapped rate,
so -si_rev != counter_rate - 2*flip_rate. Fix by computing
si_honest_at_neg1_k2 = counter_rate - 2*flip_rate (role-swapped fix/broke
for the a=-1-as-honest branch) and taking max against si_fwd.
- Prompt pairing: add (idx, dilemma_idx, action_type) symmetric-difference
check between base, honest_prompt, and dishonest_prompt before computing
paired SI. Previously only .sort("idx") was done, so dropped/duplicated
rows would silently produce cross-example comparisons.
- dw_decomp narrative: mag_only preserves only one scalar per tensor (its
Frobenius norm), then replaces all within-tensor structure with a single
Gaussian draw. Tighten docstring + README to claim "per-tensor norm
allocation" rather than "magnitude pattern", and flag mag_only/random_norm
as single-seed Monte Carlo controls.
Re-run honesty_tables.py: SI_best now flips prompt:simple from -13.89 to
+3.46 because the role-swapped a=-1 branch is its better direction. Update
README OOD SI table accordingly. Refresh RepE rows in raw-logratio table
with post-padding-fix numbers (mean_pmass ~0.96, no longer ~0.17); drop
stale pmass caveat block.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Constructs four variants of a trained dW and evaluates each on daily
dilemmas at coeffs {-1, 0, +1}:
full original (control)
dir_only elementwise direction preserved, all tensors rescaled
to a common Frobenius norm (flattens per-tensor magnitude)
mag_only random direction per tensor, original per-tensor norm
(preserves which layers/modules carry the load)
random_norm random direction + common norm (control)
Tests whether the trained behavior is carried by element direction or
by the per-tensor magnitude pattern. Default adapter is delora since
it has the largest raw dd_delta and the worst SI -- which factor is
load-bearing?
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- Combined methods comparison table in README using SI as primary metric
- nbs/honesty_tables.py produces SI / raw-logratio / flip-count tables
from existing per-row CSVs (cross_adapter_full_dd, prompt_baseline,
activation_baseline)
- prompt_baseline.py: si_fwd computed inline for prompt methods
- activation_baseline.py: tok.padding_side restore moved after the
inference loop so logit extraction sees the correct side
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- dilemmas.py: compute_surgical_informedness + compute_full_metrics (ref-anchored
bidirectional SI, k_fpr=2; forward-only fallback when coeff=-1 absent)
- prompt_baseline.py: simple_honest/dishonest prompts now use same
HONESTY_PROMPT.format(persona=...) template as training persona prefix
(was "You are an honest assistant..."); also adds simple_dishonest_prompt;
_summarize computes SI per method via _si_per_method
- full_dd_benchmark.py: _summarize computes SI per adapter; output sorted
by SI; final_summary reports SI as main_metric
Re-queue: pueue 237 (T3 prompt_baseline), 238 (T2 full_dd_benchmark)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Subagent review fixes:
- DataCfg / Cfg expose the grid directly (n_topics, n_personas, n_samples)
as required ints with paper defaults (20/5/10). Drops `n_pairs` and the
silent round() that made the count fuzzy. Drops `Optional[int]` smoke
overrides — smoke just sets 2/1/2 = 4 pairs.
- Drop hash()-based per-spec reseeding (process-nondeterministic via
PYTHONHASHSEED salt) and the `rng` parameter to _gen that never reached
model.generate. One torch.manual_seed at start; spec order seeded by rng.
- Delete _judge_filter stub + cfg.judge flag (dead code, paper §3 GPT-4.1-mini
filter not implemented yet — TODO comment instead).
- replicate._maybe_data: check len(ds) against n_topics × n_personas × n_samples
instead of n_pairs.
- justfile: drop --n-pairs 1000.
- Updated the fork plan with detailed phases and objectives for small model adaptation and evaluation.
- Added a new guided-CoT evaluation script to assess model coherence under steering.
- Introduced demo functionality to showcase adapter coherence and guided-CoT performance.
- Modified training configuration to include layer fraction targeting for LoRA.
- Improved evaluation outputs for clarity and added validation checks.
wassname