Metraipolis › archive › dc_ae59df8324e5

J-Zero: Unified Challenger--Solver--Judge Self-Evolution from Zero Data — arXiv:2608.26582

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.26582
License (as archived)
CC0-1.0
Archived
2026-10-17T14:27:53.413Z
SHA-256 of the archived text
ea005965f475657a1f7cc6efed8191f99c7787295469c7c667afc2460d852dc0
Stable URL of this page
https://metraipolis.com/documento/dc_ae59df8324e5

How to cite

J-Zero: Unified Challenger--Solver--Judge Self-Evolution from Zero Data — arXiv:2608.26582 — arXiv (en), https://arxiv.org/abs/2608.26582 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ae59df8324e5 (Metraipolis archive dc_ae59df8324e5, sha256 ea005965f475657a1f7cc6efed8191f99c7787295469c7c667afc2460d852dc0)
BibTeX
@misc{dcae59df8324e5,
  title = {J-Zero: Unified Challenger--Solver--Judge Self-Evolution from Zero Data — arXiv:2608.26582},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.26582},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ae59df8324e5; sha256 ea005965f475657a1f7cc6efed8191f99c7787295469c7c667afc2460d852dc0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.26582. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Self-evolving language models have recently emerged as a promising path toward superintelligence, with the advantage of reducing the cost of human supervision. While considerable progress has been made in verifiable domains, self-evolution in unverifiable domains remains less explored. We propose Judge co-adaptation from Zero data (J-Zero), a unified Challenger--Solver--Judge self-evolution framework that supports self-improvement across both domains. The Challenger and Solver co-evolve through an adversarial interaction: the Challenger generates increasingly difficult tasks, while the Solver learns to produce higher-quality responses to them. In parallel, the Judge co-adapts using preference pairs whose ordering is known in advance from how each response was produced, i.e., the Solver's answer over the Challenger's, and the Solver's decomposed-and-recombined answer over its one-shot answer, rather than from the Judge's own scores. J-Zero outperforms the baselines by an average of 4.2 points on verifiable and 8.0 points on unverifiable domains, and continues to improve through at least ten iterations, whereas the baselines degrade after two. Further analysis identifies Judge co-adaptation as the key driver of this sustained improvement.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-28 in cs.AI).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.26582 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T13:16:53.647Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).