Card: document willow_real_evalset / willow_synth_fit_evalset long-round campaigns

Add evalset rungs: real Willow + fit-rung extended long-round (r110-r250) decoder results + outcomes

Card: real outcomes now fully consistent (1790/1790); note per-decoder shot counts

Real beliefmatching outcomes: replace 50k-shot with the 10k-shot run matching the published CSV (now fully consistent)

Card: all five rungs now ship per-shot outcomes; document real beliefmatching 50k-vs-10k nuance

Add per-shot outcomes for real Willow + syndrome rungs (pulled from Gautschi scratch)

Card: drop ambiguous sensitivity viewer config, document dir-name-carries-param gotcha

Re-store npz/pt/png as plain blobs (no LFS)

Stop LFS-tracking npz/pt/png (all files <4MB, keep plain git clone working)

Add full Ising-sim2real experiment results: ladder, sensitivity, bootstrap, 2x2, checkpoints

initial commit