Imagen de Portada

The Long-Horizon Task Mirage? Diagnosing Where and Why Agentic Systems Break

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Wang, Xinyu Jessica, Bai, Haoyue, Sun, Yiyou, Wang, Haorui, Zhang, Shuibai, Hu, Wenjie, Schroder, Mya, Mutlu, Bilge, Song, Dawn, Nowak, Robert D
Formato:	Preprint
Publicado:	2026
Materias:	Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

How and Why LLMs Generalize: A Fine-Grained Analysis of LLM Reasoning from Cognitive Behaviors to Low-Level Patterns
por: Bai, Haoyue, et al.
Publicado: (2025)

Where's the liability in the Generative Era? Recovery-based Black-Box Detection of AI-Generated Content
por: Bai, Haoyue, et al.
Publicado: (2025)

LearnMate: Enhancing Online Education with LLM-Powered Personalized Learning Plans and Support
por: Wang, Xinyu Jessica, et al.
Publicado: (2025)

Climbing the Ladder of Reasoning: What LLMs Can-and Still Can't-Solve after SFT?
por: Sun, Yiyou, et al.
Publicado: (2025)

LearnMate^2: Design and Evaluation of an LLM-powered Personalized and Adaptive Support System for Online Learning
por: Wang, Xinyu Jessica, et al.
Publicado: (2026)

RL Grokking Recipe: How Does RL Unlock and Transfer New Algorithms in LLMs?
por: Sun, Yiyou, et al.
Publicado: (2025)

MIRAGE-Bench: LLM Agent is Hallucinating and Where to Find Them
por: Zhang, Weichen, et al.
Publicado: (2025)

Don't Break the Cache: An Evaluation of Prompt Caching for Long-Horizon Agentic Tasks
por: Lumer, Elias, et al.
Publicado: (2026)

Where we left off Crowned Coil LLC
por: Brown, Mya
Publicado: (2026)

Why and How LLMs Hallucinate: Connecting the Dots with Subsequence Associations
por: Sun, Yiyou, et al.
Publicado: (2025)

VeriPlan: Integrating Formal Verification and LLMs into End-User Planning
por: Lee, Christine, et al.
Publicado: (2025)

Agentic Aggregation for Parallel Scaling of Long-Horizon Agentic Tasks
por: Lee, Yoonsang, et al.
Publicado: (2026)

Toward Agentic AI: Task-Oriented Communication for Hierarchical Planning of Long-Horizon Tasks
por: Huang, Sin-Yu, et al.
Publicado: (2026)

U-Define: Designing User Workflows for Hard and Soft Constraints in LLM-Based Planning
por: Lee, Christine P, et al.
Publicado: (2026)

RoboClaw: An Agentic Framework for Scalable Long-Horizon Robotic Tasks
por: Li, Ruiying, et al.
Publicado: (2026)

Can Aha Moments Be Fake? Towards Quantifying Decorative and True Thinking in Chain-of-Thought
por: Zhao, Jiachen, et al.
Publicado: (2025)

AHA: Human-Assisted Out-of-Distribution Generalization and Detection
por: Bai, Haoyue, et al.
Publicado: (2024)

Toward Family-Robot Interactions: A Family-Centered Framework in HRI
por: Cagiltay, Bengisu, et al.
Publicado: (2024)

Hierarchy-of-Groups Policy Optimization for Long-Horizon Agentic Tasks
por: He, Shuo, et al.
Publicado: (2026)

Sci-VLA: Agentic VLA Inference Plugin for Long-Horizon Tasks in Scientific Experiments
por: Pang, Yiwen, et al.
Publicado: (2026)

Memory as Action: Autonomous Context Curation for Long-Horizon Agentic Tasks
por: Zhang, Yuxiang, et al.
Publicado: (2025)

Modular Jets for Supervised Pipelines: Diagnosing Mirage vs Identifiability
por: Sanyal, Suman
Publicado: (2025)

Crowdsourcing Task Traces for Service Robotics
por: Porfirio, David, et al.
Publicado: (2024)

Exploring the Use of Robots for Diary Studies
por: Xu, Michael F., et al.
Publicado: (2025)

Toward Ultra-Long-Horizon Agentic Science: Cognitive Accumulation for Machine Learning Engineering
por: Zhu, Xinyu, et al.
Publicado: (2026)

Diagnosing Hate Speech Classification: Where Do Humans and Machines Disagree, and Why?
por: Yang, Xilin
Publicado: (2024)

Deep Active Learning in the Open World
por: Xie, Tian, et al.
Publicado: (2024)

The Mirage of Performance Gains: Why Contrastive Decoding Fails to Mitigate Object Hallucinations in MLLMs?
por: Yin, Hao, et al.
Publicado: (2025)

MemWeaver: Weaving Hybrid Memories for Traceable Long-Horizon Agentic Reasoning
por: Ye, Juexiang, et al.
Publicado: (2026)

MAP: Multi-user Personalization with Collaborative LLM-powered Agents
por: Lee, Christine, et al.
Publicado: (2025)

Sprout: Designing Expressivity for Robots Using Fiber-Embedded Actuator
por: Koike, Amy, et al.
Publicado: (2024)

Factors that Affect Personalization of Robots for Older Adults
por: Stegner, Laura, et al.
Publicado: (2024)

Tangible Scenography as a Holistic Design Method for Human-Robot Interaction
por: Koike, Amy, et al.
Publicado: (2024)

Towards Robust Out-of-Distribution Generalization: Data Augmentation and Neural Architecture Search Approaches
por: Bai, Haoyue
Publicado: (2024)

Strategy Executability in Mathematical Reasoning: Leveraging Human-Model Differences for Effective Guidance
por: Liang, Weida, et al.
Publicado: (2026)

Mirage or Method? How Model-Task Alignment Induces Divergent RL Conclusions
por: Wu, Haoze, et al.
Publicado: (2025)

Agentic Feature Augmentation: Unifying Selection and Generation with Teaming, Planning, and Memories
por: Gong, Nanxu, et al.
Publicado: (2025)

Laser: Governing Long-Horizon Agentic Search via Structured Protocol and Context Register
por: Wang, Shuting, et al.
Publicado: (2025)

Search More, Think Less: Rethinking Long-Horizon Agentic Search for Efficiency and Generalization
por: Chen, Qianben, et al.
Publicado: (2026)

LEAD: Breaking the No-Recovery Bottleneck in Long-Horizon Reasoning
por: Pushkin, Denys, et al.
Publicado: (2026)