Newest papers

AnyGoal: On World Model Generalization Beyond Expert Rollouts

Heejeong Nam, Divyat Mahajan, Quentin Le Lidec, Randall Balestriero, David Alexander Klindt · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.26844 · cited by 0 works (OpenAlex) · open-access PDF

Indonesian

Tujuan dari model dunia adalah untuk menangkap dinamika lingkungan. Ini memiliki daya tarik yang besar yang memungkinkan agen untuk merencanakan dan mensimulasikan jalur tindakan dengan bantuan model dunia. Untuk penggunaan ini dapat diandalkan, bagaimanapun, dinamika yang dipelajari harus berperilaku jauh melampaui kombinasi keadaan-aksi spesifik yang diamati dalam peluncuran ahli. Terlepas dari ini, hubungan antara cakupan data, generalisasi dinamika, dan kinerja perencanaan downstream tetap belum dipelajari. Kami mempelajari hubungan ini melalui lensa generalisasi komposisi: dapatkah model dunia mengkombinasikan faktor-faktor keadaan dan tindakan yang akrab menjadi dinamika yang valid? Untuk menjawab ini, kami memperkenalkan AnyGoal, benchmark varian studi tugas yang ditransfer tujuan, yang dibangun dari lingkungan model dunia yang digunakan secara luas, bersama dengan perpustakaan perenc

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

Abstract

The goal of a world model is to capture the dynamics of an environment. This has the great appeal that it allows agents to plan and simulate a course of actions with the help of a world model. For these uses to be reliable, however, the learned dynamics must behave well beyond the specific state-action combinations observed in expert rollouts. Despite this, the relationship between data coverage, dynamics generalization, and downstream planning performance remains underexplored. We study this relationship through the lens of compositional generalization: can a world model recombine familiar state and action factors into valid dynamics? To answer this, we introduce AnyGoal, a benchmark of goal-shifted task variants built from widely used world-model environments, together with an open-source library. Each task is designed to be within reach of a world model that has learned the underlying dynamics: individual state factors and actions remain within their training support, but their combinations are underrepresented in expert trajectories. Across multiple environments and world model families, we find that current models consistently exhibit failures under these recombinations. By examining dynamics and representation failures across model families, interaction regimes, and evaluation metrics, we study where current world models generalize and where challenges remain for goal-shifted prediction and planning. Together, this framework, benchmark, and analysis provide a basis for evaluating progress toward world models that generalize to new planning tasks.

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

Topics: Computer science, Generalization, Task (project management), Action (physics), Artificial intelligence, Representation (politics)

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