docs: clarify that yacht_dice is graded on a partial sum, not the 12-round total

One-line README fix, split out of #1 so it can land on its own.

The task table currently describes `yacht_dice` as "Assign 12 dice rolls to 12 scoring categories to maximize total score". The maximization part is right, but the graded quantity is not the 12-round total — every item asks for the sum of per-round scores over a few spot-checked rounds *under* that optimal assignment:

> Using that optimal assignment, output ONLY the sum of the per-round scores for rounds 4, 7. Your final line must be exactly `Answer: <integer>` with that partial sum.

This holds for 600/600 items in the config: 2 spot-checked rounds for easy and medium, 8 for hard, in both `en` and `ko`. Someone building a scorer from the table alone would compare against the full 12-round total and score 0%.

Only the `yacht_dice` row changes — one line, no data files, no frontmatter changes.

The rest of #1 (the per-task answer-format table, the reference scorer, the `sudoku` row, and the Usage notes) stays there for separate review.

README.md

@@ -440,7 +440,7 @@ Same generators and parameters in both languages, so performance gaps isolate pr
 | `number_baseball` | Infer a hidden N-digit number from Strike/Ball hints | 100 / 100 / 100 | 100 / 100 / 100 | 600 |
 | `sat_puzzles` | Boolean satisfiability (CNF) framed as a natural-language scenario | 100 / 100 / 100 | 100 / 100 / 100 | 600 |
 | `sudoku` | 9x9 Sudoku with guaranteed-unique solutions | 100 / 100 / 100 | 100 / 100 / 100 | 600 |
-| `yacht_dice` | Assign 12 dice rolls to 12 scoring categories to maximize total score | 100 / 100 / 100 | 100 / 100 / 100 | 600 |
+| `yacht_dice` | Assign 12 dice rolls to 12 scoring categories to maximize total score; report the partial sum over spot-checked rounds | 100 / 100 / 100 | 100 / 100 / 100 | 600 |
 
 ### 2. Script Adaptations (2 tasks · EN + KO)