v8: faithful replication of the organizers' baseline (batch=1, no padding, 512 tok, naive parse, no N-forcing) + repetition_penalty=1.0

README.md

@@ -134,3 +134,17 @@ passes `repetition_penalty=1.0` explicitly.
 NFC normalisation of answers was considered and rejected: 98.15% of public golds
 are already NFC, but 13 of them are NFD-and-not-NFC, so forcing NFC would break
 those for an unmeasured gain.
+
+
+## v8 — faithful baseline replication
+
+The organizers' reference script reaches exact match **0.0729** on the hidden set
+with these exact weights. Our best is 0.0333. Before adding anything further we
+need to know whether that number is reproducible by us at all, so v8 replicates
+their script literally — trivial system prompt, no chain-of-thought, 512 new
+tokens, **batch 1 (no padding at all)**, naive line split, and **no forcing to N
+answers** — changing exactly one thing: `repetition_penalty=1.0`.
+
+If v8 lands near 0.12 the difference lives in our pipeline and can be bisected.
+If it lands near 0.08 then the reference number owes more to variance than to
+method, and the ceiling is the model.