Skip to main content
Try JUWEL

Research / Papers / Verifier-Centric Capability Growth

Paper · Methods · Map of results + PROPOSED plans

Verifier-Centric Capability Growth

A map of what worked and what failed when we added skills to a model. Plans we have not tested yet are marked PROPOSED.

Annalea Layton · Vext Labs · No date · 13 pp · ≈15 min read

  • measured map + PROPOSED program
  • toy / 1.5B-scale where stated
  • Part IV is not results
Cover · A pixel pattern made from this paper’s ID, “verifier-centric-capability-growth” · Art, not a chart
On this page · 1 section
  1. Summary

Summary

A map of what worked and what failed when we added skills to a model. Plans we have not tested yet are marked PROPOSED.

Read the full paper with the DOI link above.

Authors

Annalea Layton · Vext Labs, Inc. · ORCID 0009-0001-3695-9569

Acknowledgments

None stated in the source paper.

Cite

Layton, A. (2026). Verifier-Centric Capability Growth. Vext Labs. https://doi.org/10.5281/zenodo.21628544

@techreport{layton2026verifier,
  title  = {Verifier-Centric Capability Growth},
  author = {Layton, Annalea},
  institution = {Vext Labs},
  year   = {2026},
  type   = {Map of results + PROPOSED plans},
  doi    = {10.5281/zenodo.21628544}
}
DOI 10.5281/zenodo.21628544 CC-BY-4.0

Related

  1. 01

    Paper · 2026-07-02 · ≈25 min read

    Beyond the Scaling Ceiling

    A study of how an AI model could grow new skills in checked steps. Every result says how far it goes: tiny model, working code, or waiting on big chips.

    • workshop draft
    • PROVEN-toy
    Read the paper

  2. 02

    Paper · 2026-07-02 · ≈17 min read

    The Accretion Model

    A kind of AI model that grows by adding new parts, never by changing old ones. A stranger can check its whole history. We built a tiny version, Cultivar-0, on a regular computer.

    • BUILT/VERIFIED-CPU
    • PROVEN-toy
    Read the paper

  3. 03

    Paper · 2026-06-29 · ≈8 min read

    Sound Compounding Ratchet

    A way to keep a gain only when a checker confirms it. Plus a test for when a model thinks it passed but did not. Every win here is on tiny models only.

    • PROVEN-toy
    • $0-CPU kill-tests
    Read the paper
ShareX LinkedIn