Search results

Clarify whether slow update should be validation-gated or force-injected

Status: Open.

#22 In microsoft/SkillOpt;

· luigi-sqwish opened on May 29, 2026

Request exact splits for remaining benchmarks, especially SpreadsheetBench

Status: Open.

#21 In microsoft/SkillOpt;

· LifeIsSoSolong opened on May 29, 2026

如何解决 why？？

Status: Open.

#19 In microsoft/SkillOpt;

· 15666955178 opened on May 29, 2026

Release exact train/validation/test split manifests for reproducibility

Status: Open.

#14 In microsoft/SkillOpt;

· LifeIsSoSolong opened on May 28, 2026

SearchQA Data Split Configuration and max_completion_tokens Setting

Status: Open.

#13 In microsoft/SkillOpt;

· wenny823 opened on May 28, 2026

[Discussion] Optimizer-strength fairness in Table 1: SkillOpt uses GPT-5.5 as optimizer while several baselines (GEPA, Trace2Skill, EvoSkill) are self-contained by design

Status: Open.

#12 In microsoft/SkillOpt;

· Seanium opened on May 27, 2026

Compared to Meta Harness, will the performance gain be more significant?

Status: Open.

#11 In microsoft/SkillOpt;

· beep-bebop opened on May 27, 2026

Release optimized skills and benchmark datasets on Hugging Face

Status: Open.

#10 In microsoft/SkillOpt;

· NielsRogge opened on May 26, 2026