OSWorld-G

Computer-use grounding benchmark with original and refined instructions, released with the Jedi synthetic data pipeline, a 4-million-example Jedi dataset and Jedi-3B and Jedi-7B models.

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

desktop

Topics

evaluation, grounding

Record publication date

Not recorded

Evidence retrieved

2026-10-16T06:35:57Z

Resource ID

github:xlang-ai/osworld-g

What licenses are recorded?

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

Code

Apache-2.0 (repository LICENSE header; not a weights/data license)

Data

Not recorded

Document

Not recorded

Weights

Not recorded

Evaluation protocol

Focus

desktop GUI grounding, infeasible-target refusal

Setting

offline

Observations

screenshot, instruction

Actions / outputs

click point or box center, refusal represented by negative coordinates in the evaluator

Evaluation

The inspected Jedi runner maps predictions from resized-image coordinates back to the original image, then checks the predicted center against rectangle/polygon targets or the refusal rule. It reports overall and classification-group accuracy. This is offline grounding, not OSWorld task-execution success.

Scale and counting units

The repository retains original OSWorld-G.json and refined OSWorld-G_refined.json instructions; treat them as alternative benchmark versions, not additional disjoint tasks. Refined instructions reduce requirements beyond grounding. The same repository also contains Jedi training/rendering material, which is excluded from benchmark counts. The standalone generic eval.py entry point is a stub; the named model runner supplies image_size to its geometry helper. No task count or successful execution is inferred from repository file totals. Apache-2.0 is recorded for code only.

Compare source-reviewed profiles

Explore related topics

Related resources

• documented by: Scaling Computer-Use Grounding via User Interface Decomposition and Synthesis

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 2026-10-16T06:35:57Z. Supports: name, kind, summary, environments, topics, scope, relations, details.benchmark_protocol.

   Content fingerprint

   e814851bc3016a2abf59358f99b80afd9c2bd5894fc34baddfb7e4d0aa03bd45
2. editorial primary review

   Retrieved 2026-10-17T14:14:05Z. Supports: details.benchmark_protocol.

   Content fingerprint

   48308bba4f26b02cb7b8fbbc433bb31ec81c494dabd3af437ddbb10b8d9ce667
3. editorial primary review

   Retrieved 2026-10-17T14:22:04Z. Supports: details.benchmark_protocol.

   Content fingerprint

   64f64e0b847e0fa200aa6051e20b8035a436df00e2ef0d769688ab13fa4d3046
4. editorial primary review

   Retrieved 2026-10-17T14:21:09Z. Supports: details.benchmark_protocol.

   Content fingerprint

   b85f5685ce6c200ea648471666177859693d7f40cf50c12d08765ae68580de4a
5. license header

   Retrieved 2026-10-16T06:41:57Z. Supports: licenses.code.

   Content fingerprint

   c71d239df91726fc519c6eb72d318ec65820627232b2f796219e87dcf35d0ab4

Suggest a correction

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