Cover Image

Human-Guided Harm Recovery for Computer Use Agents

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Li, Christy, CH-Wang, Sky, Peng, Andi, Bobu, Andreea
Format:	Preprint
Published:	2026
Subjects:	Artificial Intelligence Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Learning How Hard to Think: Input-Adaptive Allocation of LM Computation
by: Damani, Mehul, et al.
Published: (2024)

Flexible Agent Alignment with Goal Inference from Open-Ended Dialog
by: Ma, Rachel, et al.
Published: (2025)

Do Androids Know They're Only Dreaming of Electric Sheep?
by: CH-Wang, Sky, et al.
Published: (2023)

Aligning Robot and Human Representations
by: Bobu, Andreea, et al.
Published: (2023)

GLIDER: Grading LLM Interactions and Decisions using Explainable Ranking
by: Deshpande, Darshan, et al.
Published: (2024)

Browsing Lost Unformed Recollections: A Benchmark for Tip-of-the-Tongue Search and Reasoning
by: CH-Wang, Sky, et al.
Published: (2025)

When Benign Inputs Lead to Severe Harms: Eliciting Unsafe Unintended Behaviors of Computer-Use Agents
by: Jones, Jaylen, et al.
Published: (2026)

AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents
by: Andriushchenko, Maksym, et al.
Published: (2024)

PluriHarms: Benchmarking the Full Spectrum of Human Judgments on AI Harm
by: Li, Jing-Jing, et al.
Published: (2026)

HarmTransform: Transforming Explicit Harmful Queries into Stealthy via Multi-Agent Debate
by: Zhu, Shenzhe
Published: (2025)

Robots That Know What to Ask: Recovering Misaligned Rewards through Targeted Explanations
by: Merker, Helena, et al.
Published: (2026)

InforME: Improving Informativeness of Abstractive Text Summarization With Informative Attention Guided by Named Entity Salience
by: Shen, Jianbin, et al.
Published: (2025)

Scaling Agents for Computer Use
by: Gonzalez-Pumariega, Gonzalo, et al.
Published: (2025)

Harm or Humor: A Multimodal, Multilingual Benchmark for Overt and Covert Harmful Humor
by: Sharshar, Ahmed, et al.
Published: (2026)

Training Computer Use Agents to Assess the Usability of Graphical User Interfaces
by: Gao, Alice, et al.
Published: (2026)

Efficient Agent Training for Computer Use
by: He, Yanheng, et al.
Published: (2025)

HarmMetric Eval: Benchmarking Metrics and Judges for LLM Harmfulness Assessment
by: Yang, Langqi, et al.
Published: (2025)

Self-HarmLLM: Can Large Language Model Harm Itself?
by: Kim, Heehwan, et al.
Published: (2025)

An Empirical Study of SFT-DPO Interaction and Parameterization in Small Language Models
by: Feng, Yuming, et al.
Published: (2026)

Test-Time Scaling with Diffusion Language Models via Reward-Guided Stitching
by: Miles, Roy, et al.
Published: (2026)

VideoAgentTrek: Computer Use Pretraining from Unlabeled Videos
by: Lu, Dunjie, et al.
Published: (2025)

MMedAgent: Learning to Use Medical Tools with Multi-modal Agent
by: Li, Binxu, et al.
Published: (2024)

ChatSOP: An SOP-Guided MCTS Planning Framework for Controllable LLM Dialogue Agents
by: Li, Zhigen, et al.
Published: (2024)

Survival at Any Cost? LLMs and the Choice Between Self-Preservation and Human Harm
by: Mohamadi, Alireza, et al.
Published: (2025)

Masked IRL: LLM-Guided Reward Disambiguation from Demonstrations and Language
by: Hwang, Minyoung, et al.
Published: (2025)

Stress-Testing Model Specs Reveals Character Differences among Language Models
by: Zhang, Jifan, et al.
Published: (2025)

StealthGraph: Exposing Domain-Specific Risks in LLMs through Knowledge-Graph-Guided Harmful Prompt Generation
by: Zheng, Huawei, et al.
Published: (2026)

Booster: Tackling Harmful Fine-tuning for Large Language Models via Attenuating Harmful Perturbation
by: Huang, Tiansheng, et al.
Published: (2024)

Towards Comprehensive Detection of Chinese Harmful Memes
by: Lu, Junyu, et al.
Published: (2024)

Agent S: An Open Agentic Framework that Uses Computers Like a Human
by: Agashe, Saaket, et al.
Published: (2024)

`For Argument's Sake, Show Me How to Harm Myself!': Jailbreaking LLMs in Suicide and Self-Harm Contexts
by: Schoene, Annika M, et al.
Published: (2025)

Preference-Conditioned Language-Guided Abstraction
by: Peng, Andi, et al.
Published: (2024)

Agent S2: A Compositional Generalist-Specialist Framework for Computer Use Agents
by: Agashe, Saaket, et al.
Published: (2025)

Milestone-Guided Policy Learning for Long-Horizon Language Agents
by: Wang, Zixuan, et al.
Published: (2026)

QuickLAP: Quick Language-Action Preference Learning for Semi-Autonomous Agents
by: Nader, Jordan Abi, et al.
Published: (2025)

FutureWeaver: Planning Test-Time Compute for Multi-Agent Systems with Modularized Collaboration
by: Jung, Dongwon, et al.
Published: (2025)

HALF: Harm-Aware LLM Fairness Evaluation Aligned with Deployment
by: Mekky, Ali, et al.
Published: (2025)

TUMIX: Multi-Agent Test-Time Scaling with Tool-Use Mixture
by: Chen, Yongchao, et al.
Published: (2025)

Structsum Generation for Faster Text Comprehension
by: Jain, Parag, et al.
Published: (2024)

SyncMind: Measuring Agent Out-of-Sync Recovery in Collaborative Software Engineering
by: Guo, Xuehang, et al.
Published: (2025)