Metraipolis › archive › dc_f95c677df24f

Scaling Large Reasoning Models beyond Human Supervision: A Path toward Superintelligence — arXiv:2608.31075

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.31075
License (as archived)
CC0-1.0
Archived
2026-10-17T07:46:59.157Z
SHA-256 of the archived text
c328192fa76295496ea407ab484d14fb897df8d11b1ee8c6e40ee603fd524ab5
Stable URL of this page
https://metraipolis.com/documento/dc_f95c677df24f

How to cite

Scaling Large Reasoning Models beyond Human Supervision: A Path toward Superintelligence — arXiv:2608.31075 — arXiv (en), https://arxiv.org/abs/2608.31075 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f95c677df24f (Metraipolis archive dc_f95c677df24f, sha256 c328192fa76295496ea407ab484d14fb897df8d11b1ee8c6e40ee603fd524ab5)
BibTeX
@misc{dcf95c677df24f,
  title = {Scaling Large Reasoning Models beyond Human Supervision: A Path toward Superintelligence — arXiv:2608.31075},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.31075},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f95c677df24f; sha256 c328192fa76295496ea407ab484d14fb897df8d11b1ee8c6e40ee603fd524ab5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.31075. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Recent advances in large reasoning models (LRMs) have shown that reinforcement learning with verifiable rewards (RLVR) can substantially improve reasoning in mathematics and code, where outcomes can be checked automatically. Extending this progress to open-ended and agentic tasks remains difficult because reliable rewards are harder to obtain and direct human supervision cannot keep pace with the scale and complexity of model-generated experience. This paper studies how LRMs can continue to improve as human supervision gradually recedes from the learning loop. We examine two connected dimensions of this problem. The reward axis traces the development from per-instance human judgments to reusable verifiers and rewards that operate even without human feedback. The experience axis examines how learning can progress from human-curated tasks and environments toward self-generated curricula, constructed environments, and autonomous co-evolution. We connect these dimensions through a five-level ladder from L0 to L4 that identifies which parts of the learning process remain under continued human control. Our analysis further highlights the risks introduced by increasingly autonomous rewards and experience generation, including reward hacking, feedback drift, curriculum collapse, and environment errors. Consequently, we also provide the evaluation around three complementary objects: policy capability, feedback fidelity, and experience quality. This analysis provides a structured account of current approaches to scaling LRMs beyond human supervision and the open problems involved in developing self-sustaining learning systems toward superintelligence. Furthermore, we maintain a continuously updated \href{ this https URL }{GitHub repository} to track the latest advances.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.31075 — 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 2026-10-17T07:28:26.638Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).