# Implementation task tracker — WP-AI-0001

Legend: [x] done · [~] in progress · [ ] todo · (gate) quality gate per CLAUDE.md §10

# Sprint 0 — Research audit

Locate UQO Research workspace, conventions, remotes
Environment audit (laptop + cluster) → docs/ENVIRONMENT_AUDIT.json
Literature verification → literature/
RESEARCH_GAP, PREREGISTRATION v0.1, EXPERIMENT_PLAN, COMPUTE_BUDGET, RISK_REGISTER, REPRODUCIBILITY, logs
Project skeleton, pyproject, tests (Gate A)
scripts/build_figures.py, scripts/build_paper.sh, paper skeleton compiles (placeholders)
git repo + origin = gitsrv:srv/git/wpai0001_uqo.git

OPSEQ / FSA / MORPH generators + exact validators + disjointness + hashing
Toy transformer with hook points; no-op tests (Gate C part 1)
Training loop (teacher-forced scratchpad), greedy decoding, metrics, cluster bootstrap
Pilot run (3 seeds) — gap qualification (Gate B passed; run pilot_tier0_opseq_451ef5087182)
FSA / MORPH toy baselines
trace_stride = 2 difficulty variant baseline

Within-donor localization (patching, ablation, heads) with sham baseline and reproducibility (RQ1 pilot: Ω executed at the read position; L0 attention + L0–L2 MLPs)
Add embed/resid_pre family and path patching Ω-token → r1 to the sweep
Subspace localization (DAS-style / PCA of Ω-contrast) with norm-matched random subspaces
Necessity test on donor (H3) with control-stratum retention

Alignment module: identity, Procrustes, ridge, CCA, random orthogonal; task-row vs task-free fits; leakage audit by item hash
Activation graft + sham + norm-matched random + shuffled-token/layer + unrelated-donor controls
Gate D: reproducibility across 9 pairs, overhead disclosed (donor in the loop), map parameter counts

Subspace, steering, adapter, weight, persistent grafts; supervised-adapter and distillation references
Layer grid, dose, width/depth mismatch, stride-2; ablation → rescue; reverse graft; rule-vs-cache
Confirmatory freeze (PREREGISTRATION v1.0, D10) and confirmatory run + analysis (H2, H5 supported; H4 not)
Tier-1 MLX layer wrapping with bitwise no-op test; task-vector localization + transfer script
[~] Tier-1 development run (4 pairs) → Tier-1 freeze → confirmatory split run
[~] Tier-2: Qwen2.5-7B ↔ 3B pairs on M3U96b
Transformers equivalence test for the MLX backend (not done; reported as a limitation)

[~] Paper sections written for Tier 0; Tier 1 pending; claim-evidence table (Gate F) pending
Internal multi-agent review (UQO-WP-REVIEW pipeline)