analiza aiqbench
ReHoPER introduces a zero-shot, inference-only framework that enhances large language model reasoning by iteratively planning and answering intermediate questions across multiple paths. This task-agnostic approach demonstrates significant performance gains on compositional reasoning benchmarks without requiring labeled data or specific prompt engineering. The release of the implementation and iLLC benchmark provides a valuable resource for advancing agentic planning and complex problem-solving capabilities in AI systems.
ReHoPER: Receding-Horizon Planning for Enhanced Reasoning
arXiv:2610.00940v1 Announce Type: cross Abstract: We propose ReHoPER, an inference-only, zero-shot method that improves large language models' reasoning by generating and answering intermediate questions along multiple paths before the final answer. It iteratively plans a horizon of candidate intermediate questions, selects one to answer, and replans from the updated history. ReHoPER is task-agnostic, using the same generic instructions across datasets and models without labeled data or task-specific prompt design. Across multiple datasets, including iLLC, a new controlled benchmark for compositional reasoning, ReHoPER outperforms strong baselines, with the largest gains in the most compositional settings. Our implementation and the iLLC generator are publicly available to support future work.
Abstract:We propose ReHoPER, an inference-only, zero-shot method that improves large language models' reasoning by generating and answering intermediate questions along multiple paths before the final answer. It iteratively plans a horizon of candidate intermediate questions, selects one to answer, and replans from the updated history. ReHoPER is task-agnostic, using the same generic instructions across datasets and models without labeled data or task-specific prompt design. Across multiple datasets, including iLLC, a new controlled benchmark for compositional reasoning, ReHoPER outperforms strong baselines, with the largest gains in the most compositional settings. Our implementation and the iLLC generator are publicly available to support future work.
Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
arXiv:2610.00940 [cs.CL]
(or arXiv:2610.00940v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2610.00940 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 203 cuvinte
Subiecte conexe
Any House Any Task: Scalable Long-Horizon Planning for Abstract Human Tasks
arXiv · 4 aug. 2026
Toward Unified Robot Learning: Bridging Representation, Vision-Language-Action, and World Models
arXiv · 4 sept. 2026
SEDualVLN: A Spatially-Enhanced Dual-System for Vision-Language Navigation
arXiv · 5 iun. 2026
Receding-Horizon Next-Best-View Planner for Autonomous Leaf Surface Reconstruction
arXiv · 3 aug. 2026