{"ID":5937825,"CreatedAt":"2026-07-07T03:14:33.014478982Z","UpdatedAt":"2026-07-09T07:52:46.28543944Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.04542","arxiv_id":"2607.04542","title":"Auto: The AGI Compiler","abstract":"Every LLM agent run re-derives its behavior token by token on a frontier model: brilliant, expensive, slow, and unbounded. We present Auto, a compiler that records live agent behavior, measures which parts are secretly deterministic, extracts them into verified programs or distilled specialists, and emits cognition binaries: WebAssembly artifacts whose manifests carry measured guarantees and whose declared capabilities are physically enforced by the sandbox. A tiered runtime executes compiled behavior behind conformally calibrated guards; guard trips deopt to the reference agent, and the captured trace recompiles back down, so nothing is figured out twice. We use \"AGI compiler\" in one narrow, testable sense: a system that autonomously converts novel experience into permanent, verified, near-free skill while measuring what it does not know. On AUTO-BENCH, a benchmark we introduce and pre-register, 87.1% of 560 recorded frontier-agent spans are witnessed-deterministic (three of the four censused task families measure 100.0%). On a 300-item stream with three scheduled distribution shifts, the closed loop compiles three artifact generations and drives marginal cost from 59 to 2 micro-dollars per item (6.4x end-to-end) at 96.9% parity on witnessed inputs with zero errors. The same stream also quantifies the failure modes: a loose guard silently mislabels 48.9% of compiled answers, and an unfaithful deopt reference causes the verification gate to refuse recompilation. Calibration and reference fidelity, not model capability, decide whether cheap stays correct. Code: https://github.com/RightNow-AI/auto","short_abstract":"Every LLM agent run re-derives its behavior token by token on a frontier model: brilliant, expensive, slow, and unbounded. We present Auto, a compiler that records live agent behavior, measures which parts are secretly deterministic, extracts them into verified programs or distilled specialists, and emits cognition bin...","url_abs":"https://arxiv.org/abs/2607.04542","url_pdf":"https://arxiv.org/pdf/2607.04542v1","authors":"[\"Jaber Jaber\",\"Osama Jaber\"]","published":"2026-07-05T23:09:24Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AI\",\"cs.SE\"]","methods":"[\"Large Language Model\"]","has_code":false,"code_links":[{"ID":613990,"CreatedAt":"2026-07-07T03:14:33.014478982Z","UpdatedAt":"2026-07-07T03:14:33.014478982Z","DeletedAt":null,"paper_id":5937825,"paper_url":"https://arxiv.org/abs/2607.04542","paper_title":"Auto: The AGI Compiler","repo_url":"https://github.com/RightNow-AI/auto","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}