De Wang, Peng Qi · 2026 · arXiv
DOI: 10.48550/arXiv.2610.27465 · open-access PDF
Platform agen mulai meningkatkan agen mereka sendiri: mereka membaca jejak eksekusi agen yang didistribusikan, menghasilkan perubahan, memiliki verifier otomatis menyetujui itu, dan melepaskannya tanpa penulis manusia. Berapa banyak loop ini dapat berjalan dengan aman tanpa manusia? Kami menyarankan dua jumlah yang dapat diukur oleh platform. Verifier strength $σ$ adalah seberapa baik verifier yang tersedia - sering hakim LLM - setuju dengan verifier referensi yang dapat diandalkan; itu membatasi seberapa banyak langkah dapat meningkatkan agen. Recoverability adalah seberapa murah perubahan buruk dapat dibatalkan: apakah membatalkannya membutuhkan persetujuan verifier baru, dan seberapa banyak t
Machine translation (Meta M2M100); the English original follows.
Agent platforms are starting to improve their own agents: they read a deployed agent's execution traces, derive a change, have an automated verifier approve it, and release it with no human author. How much of this loop can safely run without a human? We propose two quantities a platform can measure. Verifier strength $σ$ is how well the available verifier -- often an LLM judge -- agrees with a trusted reference verifier; it limits how much a step can improve the agent. Recoverability is how cheaply a bad change can be undone: whether undoing it needs a new verifier approval, and how much of t
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.27465, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange