Alignment in Time: Peak-Aware Orchestration for Long-Horizon Agentic Systems

TL;DR

This paper introduces APEMO, a runtime scheduling method that improves long-horizon autonomous systems by dynamically managing computational resources based on temporal-affective signals, enhancing reliability and trajectory quality.

Contribution

It presents APEMO, a novel temporal-affective control framework that optimizes resource allocation without modifying models, focusing on critical trajectory segments for better alignment.

Findings

APEMO improves trajectory-level quality in multi-agent simulations.

APEMO increases reuse probability in LLM-based workflows.

Temporal control framing enhances long-horizon agent reliability.

Abstract

Traditional AI alignment primarily focuses on individual model outputs; however, autonomous agents in long-horizon workflows require sustained reliability across entire interaction trajectories. We introduce APEMO (Affect-aware Peak-End Modulation for Orchestration), a runtime scheduling layer that optimizes computational allocation under fixed budgets by operationalizing temporal-affective signals. Instead of modifying model weights, APEMO detects trajectory instability through behavioral proxies and targets repairs at critical segments, such as peak moments and endings. Evaluation across multi-agent simulations and LLM-based planner--executor flows demonstrates that APEMO consistently enhances trajectory-level quality and reuse probability over structural orchestrators. Our results reframe alignment as a temporal control problem, offering a resilient engineering pathway for the development of long-horizon agentic systems.