This note records the current isolated secp256k1 proof-aware search surface.
It is intentionally separate from the audited direct P-256 leaderboard.
- This is a curve-specific experimental track, not an update to the P-256 result.
- The scaffold reuses the direct 256-bit accounting structure so that
secp256k1-specific reduction and GLV-style choices can be audited in a narrow search surface. - The only new curve-specific families in this pass are:
special_prime_strategyendomorphism_strategy
- full experimental winner:
1137logical qubits - Toffolis per run:
840,664,542,458 - depth:
838,004,479,352 - runs:
22 - paper-explicit floor on the isolated scaffold:
1193 - best paper-motivated floor:
1155 - full surface size:
9configurations
Current winner:
special_prime_strategy = reduction_focusedendomorphism_strategy = glv_shared_controlrun_model = reported_22_runs
Best paper-motivated point:
special_prime_strategy = constant_heavyendomorphism_strategy = glv_balancedrun_model = reported_22_runs
The current experimental winner uses exactly 2 repo-engineered substitutions:
special_prime_strategy = reduction_focusedendomorphism_strategy = glv_shared_control
The best key at each repo-engineered count is:
0 -> (1155, 20,159,291,478,568, 916,838,492,872)1 -> (1145, 19,093,392,190,004, 865,330,712,374)2 -> (1137, 18,494,619,934,076, 838,004,479,352)
For the current experimental winner:
- peak stage:
point_addition_peak tree_storage_stage = 942point_addition_peak = 1137legendre_compression_peak = 886
Peak-stage provenance totals:
paper-explicit = 1193repo-engineered = -56
- direct materialization and independent-counter agreement for the default scaffold point
- direct materialization and independent-counter agreement for the current experimental winner
- exhaustive search over all
9configurations in the current surface - proof-search reproduction from the default seed and stability from the current-winner seed
- liveness audit and provenance-gap regression coverage
- correctness-backed
secp256k1packing, reduction, and compression checks
python3 -m ecdsa_autoresearch.cli audit-chevignard-secp256k1-liveness --use-current-winner
python3 -m ecdsa_autoresearch.cli audit-chevignard-secp256k1-provenance-gap
python3 -m ecdsa_autoresearch.cli run-chevignard-secp256k1-proof-search \
--initial-config current-winner \
--rounds 4 \
--max-changes-per-trial 2 \
--no-build --no-cache