Lock alpha scope, tracker auth, TOPLOC fraud verification, and deferred multi-tracker money-path work; supersede legacy fraud issues with ADR-0018. Co-authored-by: Cursor <cursoragent@cursor.com>
2.2 KiB
2.2 KiB
Status: ready-for-agent
06 — FRAUD: TOPLOC integration (teacher-forced audit primitive)
What to build
Adopt TOPLOC (MIT, pip install toploc) for activation fingerprint commit and verify. Replace string-equality validator checks with teacher-forced prefill + TOPLOC tolerance matching.
Estimated effort: 2+ sessions — split into subtasks below without separate issue files.
| Subtask | Owner package | Deliverable |
|---|---|---|
| Validator audit primitive | packages/validator/ |
Teacher-forced prefill, TOPLOC verify, unit tests with stub tensors |
| Node runtime commitments | packages/node/ (if prover-side) |
On-demand activation fingerprint generation on audit-selected requests |
Code refs:
packages/validator/meshnet_validator/__init__.py—_run_reference,_outputs_match(~92–148)packages/validator/README.md— deterrence math (update for 19× at p=0.05)- Research:
.scratch/alpha-hardening/research-verifiable-inference.md§8 layers 1–2, build-vs-adopt table
Pin one canonical precision/quantization per model preset. Add toploc to validator (and node if prover-side) dependencies.
Test-first
- Red: validator compares final text strings — fails on cross-GPU honest divergence (document expected).
- Green: stub activation tensors + TOPLOC proofs round-trip in unit test.
- Integration: reference node teacher-forces tokens; verify accepts honest proof, rejects swapped precision.
Acceptance criteria
toplocdependency declared;build_proofs_*/verify_proofs_*wired- Validator re-runs claimed token sequence as prefill, not free generation
- Model preset documents canonical dtype/quantization
- README updated: 19× deterrence at 5% audit (research §1.1)
- Tests with deterministic stub tensors (no GPU required in CI)
ADR links
- ADR-0018 §2
- Research: research-verifiable-inference.md §8, §9 build-vs-adopt
Blocked by
05-a1-a5-persist-strike-ban-reputation.md
Prod gate: do not enable production audit thresholds until 21-honest-noise-calibration-corpus.md completes (see README Phase 2 note).