Long-Horizon Agent Tasks

Follow this topic Ask about this topic

This field studies agents that must carry goals through many dependent steps in settings such as web browsing, computer use, software engineering, and embodied environments. Common approaches include hierarchical planning and reusable skills, external or task-aware memory, context compression, tool use, verification, and reinforcement learning with milestone-based credit assignment. Benchmarks increasingly test persistent workflows, changing environments, and partial progress, but current agents still often lose task state or compound small errors as interactions grow longer. A central tension is how to extend useful reasoning and memory while controlling inference cost and ensuring that progress and final outcomes are verifiable.

Latest papers

193 papers42 papers in the last four weeks, up 83% on the four weeks before. 0.2% of all new papers.

MANSION: Multi-floor lANguage-to-3D Scene generatIOn for loNg-horizon tasks
Mar 12, 2026Lirong Che, Shuo Wen, Shan Huang +53D Scene Generation Indoor Scene Generation
Tsinghua University · AgiBot · McGill University, MILA - Quebec AI Institute
Memex(RL): Scaling Long-Horizon LLM Agents via Indexed Experience Memory
Mar 4, 2026Zhenting Wang, Huancheng Chen, Jiayun Wang +1LLM Agent Memory Memory Management for LLM Agents
Center for Advanced AI, Accenture
HiMAC: Hierarchical Macro-Micro Learning for Long-Horizon LLM Agents
Mar 1, 2026Hongbo Jin, Rongpeng Zhu, Jiayu Ding +2Long-Horizon LLM Agents Long-Horizon Agent Tasks
School of Electronic and Computer Engineering, Peking University
MemPO: Self-Memory Policy Optimization for Long-Horizon Agents
Feb 28, 2026Ruoran Li, Xinghua Zhang, Haiyang Yu +7Agent Memory Management Memory-Augmented Agents
Tsinghua University · Tongyi Lab, Alibaba Group
SWE-Hub: A Unified Production System for Scalable, Executable Software Engineering Tasks
Feb 28, 2026Yucheng Zeng, Shupeng Li, Daxiang Dong +11Synthetic Benchmark Generation Software Engineering Benchmarks
Qianfan Team, Baidu Inc.
Hierarchy-of-Groups Policy Optimization for Long-Horizon Agentic Tasks
Feb 26, 2026Shuo He, Lang Feng, Qi Wei +3Agentic RL RL for LLM Agents
Nanyang Technological University · Southeast University
KLong: Training LLM Agent for Extremely Long-horizon Tasks
Feb 19, 2026Yue LiuLLM Agent Training Long-Horizon Agent Tasks
NUS
IntentCUA: Learning Intent-level Representations for Skill Abstraction and Multi-Agent Planning in Computer-Use Agents
Feb 19, 2026Seoyoung Lee, Seobin Yoon, Seongbeen Lee +4Computer-Use Agents Long-Horizon Agent Tasks
Sookmyung Women’s University, Seoul, Republic of Korea
OdysseyArena: Benchmarking Large Language Models For Long-Horizon, Active and Inductive Interactions
Feb 5, 2026Fangzhi Xu, Hang Yan, Qiushi Sun +14AI Agent Benchmarks Long-Horizon Agent Evaluation
Xi’an Jiaotong University · The University of Hong Kong · Tsinghua University +5
Chain-of-Goals Hierarchical Policy for Long-Horizon Offline Goal-Conditioned RL
Feb 3, 2026Jinwoo Choi, Sang-Hyun Lee, Seung-Woo SeoReinforcement Learning Hierarchical RL
Department of Electrical and Computer Engineering, Seoul National University, Seoul, South Korea · Department of Automotive Engineering, Ajou University, Gyeonggi-do, South Korea
PEARL: Self-Evolving Assistant for Time Management with Reinforcement Learning
Jan 17, 2026Bingxuan Li, Jeonghwan Kim, Cheng Qian +4Long-Horizon Agent Tasks RL for LLM Agents
University of Illinois at Urbana-Champaign · Viven
MACRO-LLM: LLM-Empowered Multi-Agent Collaborative Reasoning under Spatiotemporal Partial Observability
Jan 14, 2026Handi Chen, Running Zhao, Xiuzhe Wu +2Multi-Agent LLM Systems Multi-Agent Coordination
The University of Hong Kong
Imagine-then-Plan: Agent Learning from Adaptive Lookahead with World Models
Jan 13, 2026Youwei Liu, Jian Wang, Hanlin Wang +2World Model-Based Planning Model-Based Planning
Central South University · Department of Computing, The Hong Kong Polytechnic University · College of Computer Science, Sichuan University

Mar 12, 2026J/K move · Enter open · S save

MANSION: Multi-floor lANguage-to-3D Scene generatIOn for loNg-horizon tasks

Lirong Che, Shuo Wen, Shan Huang, Chuang Wang+4

Tsinghua University · AgiBot · McGill University, MILA - Quebec AI Institute

3D Scene Generation Indoor Scene Generation Synthetic Data Generation for Robotics Long-Horizon Agent Tasks

Open paper AskPDF arXiv

Real-world robotic tasks are long-horizon and often span multiple floors, demanding rich spatial reasoning. However, existing embodied benchmarks are largely confined to single-floor in-house environments, failing to reflect the complexity of real-world tasks. We introduce MANSION, the first language-driven framework for generating building-scale, multi-floor 3D environments. Being aware of vertical structural constraints, MANSION generates realistic, navigable whole-building structures with diverse, human-friendly scenes, enabling the development and evaluation of cross-floor long-horizon tasks. Building on this framework, we release MansionWorld, a dataset of over 1,000 diverse buildings ranging from hospitals to offices, alongside a Task-Semantic Scene Editing Agent that customizes these environments using open-vocabulary commands to meet specific user needs. Benchmarking reveals that state-of-the-art agents degrade sharply in our settings, establishing MANSION as a critical testbed for the next generation of spatial reasoning and planning.