Precision CUA Bench

Evaluation resources for pixel-precise cursor grounding in code editors.

AI-assisted editorial review against a primary source. This is a catalog annotation, not an independent reproduction or endorsement.

What is this resource?

Resource type

Benchmarks

Environment

cross-platform

Topics

evaluation, grounding

Record publication date

Not recorded

Evidence retrieved

2026-10-16T06:40:52Z

Resource ID

github:microsoft/precision-cua-bench

What licenses are recorded?

Code, training data, weights, and documents can have different terms. A blank declaration means unknown, not unrestricted use.

Code

MIT (repository LICENSE header; not a weights/data license)

Data

Not recorded

Document

Not recorded

Weights

Not recorded

Evaluation protocol

Focus

pixel-precise editor cursor placement, multi-turn visual self-correction

Setting

offline

Observations

code-editor screenshot, cursor-location instruction, red-cross feedback marking previous incorrect prediction

Actions / outputs

pixel (x,y) cursor coordinate

Evaluation

The inspected evaluator scales target boxes from a 1000×1000 reference grid into image pixels and tests pixel predictions, allowing ±10 pixels around the center along any target dimension narrower than 1 pixel after scaling. It reports cumulative hit accuracy by turn (earlier hits remain correct), miss distances, any-turn hit rate, and a correction rate whose denominator is all successfully processed samples. The default budget is five turns; retain prompt/feedback variant, turn budget, and processed/skipped coverage.

Scale and counting units

The release has four editor/theme JSONL partitions (Cursor/VS Code × dark/light) and screenshot files. Partitions are not train/test splits or independent task totals. README’s initial field table calls bbox pixel coordinates, but the evaluation-format section and actual ratio_h/ratio_w=dimension/1000 code use normalized target boxes; profile follows the inspected evaluator and retains the inconsistency. Source/images were not executed or downloaded as a dataset. MIT is recorded for repository code only.

Compare source-reviewed profiles

Explore related topics

Related resources

• documented by: PrecisionCUA: Iterative Visual Refinement for Pixel-Precise Cursor Grounding in Code Editors

Primary evidence and verification

Source retrieval is separate from experiment reproduction. Use the original publication or publisher page when citing claims about the resource.

1. editorial primary review

   Retrieved 2026-10-16T06:40:52Z. Supports: name, kind, summary, environments, topics, scope, relations.

   Content fingerprint

   23d0dc8ea0a10c080cf8ae32121eade2785ae923a6b0d853d166c94c195e789b
2. editorial primary review

   Retrieved 2026-10-16T06:35:59Z. Supports: details.benchmark_protocol.

   Content fingerprint

   84ad4fb798abb776dce813cf4dab296d6e03248d8ae29388e1f06cccd27d1241
3. editorial primary review

   Retrieved 2026-10-17T14:21:08Z. Supports: details.benchmark_protocol.

   Content fingerprint

   1fda7c3980fe7a7c373478eda6ecfabff227e92c2b1f19c6fe903b408f583797
4. editorial primary review

   Retrieved 2026-10-17T14:14:06Z. Supports: details.benchmark_protocol.

   Content fingerprint

   ede84e3dda47f80fada4234caa2423db08e21ff7667a37599c03aae657b2f84b
5. editorial primary review

   Retrieved 2026-10-16T06:32:00Z. Supports: relations.

   Content fingerprint

   4d09ebfb093e43a4c44adef86c8fa8374f90d907fc33bd8e12a1315e62d5a583
6. license header

   Retrieved 2026-10-16T06:41:59Z. Supports: licenses.code.

   Content fingerprint

   c2cfccb812fe482101a8f04597dfc5a9991a6b2748266c47ac91b6a5aae15383

Suggest a correction

Open a source-backed correction. Include this record ID and the original source. Read our editorial policy and coverage boundaries.