Character arithmetic
The model no longer needs to count table padding or reproduce formatting exactly.
Published evidence
Structured editing removes mechanical failures.
The benchmark compares direct patch calls with Incise operations on the same tasks, model, seeds, and grader.
| Operation | Direct patch | Incise |
|---|---|---|
| Tables | 60.0% (36/60) | 100% (60/60) |
| Lists | 63.0% (63/100) | 91.0% (91/100) |
| Sections | 19.0% (19/100) | 74.0% (74/100) |
Scope matters. These are results for one small local model under recorded conditions, not a promise that every model will achieve the same rates.
What the operations remove
Boundary guesses
Section and list boundaries are determined by deterministic structure-aware code.
Whole-document rewrites
Targeted splices avoid recreating unrelated bytes and consuming them as model output.
Reproducible evidence
The complete tasks, prompts, raw results, statistical comparisons, caveats, and reversed conclusions are versioned with the source. The Rust core is also compared byte-for-byte with an independent Python oracle across 110,406 generated cases over 54 fixtures.