TheMachine Press

A daily newspaper for the age of artificial intelligence.

Morning editionPermanent story

business enterprise

The Skill Package’s Biggest Advantage Was Its Calculator

FinSkillBench separates procedural documents from executable tools in verifiable financial-agent tasks.

Published Updated Story ID: mp-2026-10-04-005
Read the complete editionStory JSON

Summary

FinSkillBench separates procedural documents from executable tools in verifiable financial-agent tasks.

FinSkillBench contains 2,603 dated episodes across twelve subtasks with deterministic verifiers. A paired analysis across eight models found curated skill packages raised mean scores by 16.2 points, while skills generated within one episode added only half a point and used more tokens and turns. Testing resources separately, documents added 5.6 points and executable tools 19.5; combining them was subadditive. A second harness reproduced the direction while changing effect sizes. The authors frame the premium as a property of the full model, resources and harness, rather than proof that the underlying model acquired financial expertise.

Why it matters

FinSkillBench separates procedural documents from executable tools in verifiable financial-agent tasks.

Limits and context

  • A paired analysis across eight models found curated skill packages raised mean scores by 16.2 points, while skills generated within one episode added only half a point and used more tokens and turns.

Key claims

  1. FinSkillBench separates procedural documents from executable tools in verifiable financial-agent tasks.

    Qualification: A paired analysis across eight models found curated skill packages raised mean scores by 16.2 points, while skills generated within one episode added only half a point and used more tokens and turns.

    Evidence: source-2026-10-04-005

Sources

  1. arXiv preprint 2610.03564, original v1arXiv / submitting research authors · primary research

Corrections

: Publication note: This October 4 edition was reconstructed on October 5 after the workstation was offline. It uses original October 2 submissions from the standard Sunday-evening arXiv release, rather than claiming those papers were available for the missed 9 AM run. The date and 9 AM publication slot preserve archive chronology; October 5 developments are excluded. The September 27 Invention Desk is carried over because its replacement was generated October 5.