Newest papers

Knowing When to Stop Thinking: A Dissection of Expected Free Energy as a Test-Time Compute Controller for Latent World-Model Planning

Tong Zhang, Zhiyuan Shi, Yun Peng, Tao Xie · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.28011 · cited by 0 works (OpenAlex) · open-access PDF

Russian

Планировщик, использующий модель изученного мира, может улучшить свой прогнозируемый исход при выборе худшего действия. Мы изучаем, когда ожидаемая свободная энергия (EFE) может помочь решить, сколько вычислений потратить на план. Объектив, вдохновленный EFE, сочетает расстояние целей с несогласием между стохастическими прогнозами. Мы тестируем результаты баллов кандидата, остановку и обновления онлайн-предсказателей отдельно. Остановка, когда элитные затраты стоят, уменьшает утонченность на задачах, чьи успешные платформы с глубиной, с средним успехом вблизи фиксированной глубины ссылок. При более крупных тестируемых весах, вознаграждение несогласие снижает успех с заморожен

Machine translation (Meta M2M100); the English original follows.

Abstract

A planner using a learned world model can improve its predicted outcome while choosing a worse action. We study when expected free energy (EFE) can help decide how much computation to spend on a plan. An EFE-inspired objective combines goal distance with disagreement among stochastic predictions. We test candidate scoring, stopping, and online predictor updates separately. Stopping when elite cost stalls reduces refinement on tasks whose success plateaus with depth, with mean success close to fixed-depth references. At the larger tested weights, rewarding disagreement lowers success with a frozen model. Penalizing it helps at some task and weight settings but harms others. When training data omit a region needed to reach the goal, online updates recover part of the resulting loss. The tested cost-stall rule can fail when further optimization exploits model errors, because predicted cost keeps falling after task success peaks. Cost progress can guide computation on a plateau, but protecting performance after a peak requires information about the consequences of the chosen actions.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Task (project management), Computer science, Computation, Exploit, Outcome (game theory), Energy (signal processing)

Is this paper real? This page was built from OpenAlex and arXiv records, not generated. Software can verify any reference in one call: GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.28011, 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.txt

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange