methods/Yuseon84: ED and Néel-field VMC references behind #56 (δ = 0) — scripts and outputs

#69
by Yuseon84 - opened

Scripts and outputs requested in #56, added under methods/Yuseon84/ (code MIT; reference values CC-BY-NC 4.0).

  • run_delta0.sh reproduces the two δ = 0 numbers cited in #56: 4×4 ED at n = 1, R = (2,2), P_d = 0.006197
    (equal to the published Ku-ku value), and the 6×6 Néel-field VMC at Ne = 36, R = (3,3), P_d = 0.0049 ± 0.0007
    (lowest-energy state, three seeds).
  • The rest of the package: 4×4 ED for Ne = 8–16 with per-momentum-sector values, U = 0 references, and the
    6×6 VMC table for δ ≤ 0.333 (values at δ ≥ 0.389 deliberately omitted).
  • Frozen VMC parameters in configs/, expected outputs in expected/, SHA256SUMS.txt.

This is a calibration reference, not a request to change any score of our own entries.
Prepared with AI assistance (Claude Code).

🤖 Generated with Claude Code

FINAL_Bench org

Received and accepted as a calibration/methods reference, @Yuseon84 , under your stated licenses (code MIT, values CC-BY-NC). Thank you - the frozen configs, SHA256SUMS and expected outputs are exactly what a cross-check needs. These delta=0 references (4x4 ED 0.006197, 6x6 Neel-VMC 0.0049) are being used in our review of the scoring_meta.json delta=0 A_d point; the outcome will be posted to #56/#34. No score changes are implied yet.

Ready to merge
This branch is ready to get merged automatically.

Sign up or log in to comment