aiqbench takeMachine Learning

SkillOpt demonstrates significant cross-model transferability, proving that natural-language skills optimized for one LLM can effectively enhance another without retraining the base model. This finding suggests a path toward more efficient AI deployment by decoupling skill development from specific inference engines, allowing organizations to share artifacts across different harnesses like Codex and Claude Code.

Sources coveringMarkTechPost, Latent Space

What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

Microsoft’s SkillOpt Shows Optimized Agent Skill Artifacts Transfer Across Model Scales and Between Codex and Claude Code Harnesses

MarkTechPostAugust 6, 20265 min read

Most coverage of Microsoft's SkillOpt centers on its 52/52 result. The more consequential finding is in Section 4.3: the exported best_skill.md keeps working in environments it was never trained on. A Codex-trained SpreadsheetBench skill lifted Claude Code from 22.1 to 81.8, slightly above the 80.4 that harness reached training its own skill. Retention varies sharply by task type — 102% on spreadsheets, 10% on math — which is what makes the result worth reading closely. The post Microsoft’s SkillOpt Shows Optimized Agent Skill Artifacts Transfer Across Model Scales and Between Codex and Claude Code Harnesses appeared first on MarkTechPost.

SkillOpt is a text-space optimizer developed by a team of researchers from Microsoft, Shanghai Jiao Tong University, Tongji University, and Fudan University. SkillOpt trains a single natural-language skill document while the target model stays frozen. An optimizer model reads scored rollouts and proposes bounded add/delete/replace edits. A held-out selection split accepts an edit only when the score strictly improves. The exported artifact is one file, best_skill.md. The transfer tables report three columns. Baseline is the target’s no-skill score. Direct is SkillOpt trained in-domain on that exact target. Transferred…

Excerpt shown under the publisher's RSS terms · 88 words · full text and figures are on the publisher's site

Read at MarkTechPost ↗Ask about this article