Files
oak-gpui/crates
Oleksiy Syvokon 99d7b2fa1d zeta2: Compute diff-aware chrF metric (#43485)
Zeta evals now include a character n-gram metric adapted for multi-edit diffs (“delta chrF”). It works as follows:

1. Reconstruct the original, golden (expected), and actual texts from unified diffs.
   - "original": the text before any edits
   - "golden": the text after applying the expected edits
   - "actual": the text after applying the actual edits

2. Compute n-gram count deltas between original→golden and original→actual.
   - n-grams are computed as in chrF (max n=6, whitespace ignored).

3. Compare these deltas to assess how well the actual edits match the expected edits.
   - As in standard chrF, classify n-grams as true positives, false positives, and false negatives, and report the F-beta score with beta=2.

Release Notes:

- N/A
2025-11-27 11:10:35 +02:00
..
2025-10-17 18:58:14 +00:00
2025-11-24 20:01:43 +00:00
2025-11-12 20:27:04 +00:00
2025-10-17 18:58:14 +00:00
2025-11-19 12:34:41 +01:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-11-24 13:24:26 -05:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-11-20 22:17:50 -07:00
2025-11-19 05:41:24 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-11-11 16:10:47 +01:00
2025-11-25 14:17:27 -08:00
2025-10-17 18:58:14 +00:00
2025-11-12 15:31:20 +01:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-24 07:52:51 -04:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-11-07 11:36:05 +00:00
2025-11-24 11:36:04 +02:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00
2025-10-17 18:58:14 +00:00