[2603.27905] ATLAS-RTC: Closing the Loop on LLM Agent Output with Token-Level Runtime Control

Computer Science > Machine Learning arXiv:2603.27905 (cs) [Submitted on 29 Mar 2026] Title:ATLAS-RTC: Closing the Loop on LLM Agent Output with Token-Level Runtime Control Authors:Christopher Cruz View a PDF of the paper titled ATLAS-RTC: Closing the Loop on LLM Agent Output with Token-Level Runtime Control, by Christopher Cruz View PDF HTML (experimental) Abstract:We present ATLAS-RTC, a runtime control system for autoregressive language models that enforces structured output during decoding. ATLAS-RTC monitors generation at each step, detects drift from output contracts using lightweight signals, and applies targeted interventions such as biasing, masking, and rollback. Unlike post-hoc validation or static constrained decoding, it operates in a closed loop, enabling correction before errors materialize. Across structured generation and tool-calling tasks, ATLAS-RTC improves first-attempt success rates by 20 to 37.8 percentage points, with up to 88% latency reduction in failure-dominated settings. Results show that many failures arise from decoding artifacts rather than task misunderstanding, motivating runtime control as a distinct layer in LLM systems. Subjects: Machine Learning (cs.LG) ACM classes: I.2.8 Cite as: arXiv:2603.27905 [cs.LG]   (or arXiv:2603.27905v1 [cs.LG] for this version)   https://doi.org/10.48550/arXiv.2603.27905 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Christopher Cruz [view email] [v1] Sun, 29...