Step-level Optimization for Efficient Computer-use Agents

This paper targets the cost and latency of computer-use agents that call large multimodal models at nearly every GUI step. It argues that long-horizon trajectories have heterogeneous step difficulty and studies step-level optimization so routine actions can be handled by cheaper policies while high-risk moments receive stronger compute.

• TRACE: Trajectory-robust Admission with Evidence Ordering for Efficient GUI Agents
  September 9, 2026 · arXiv
• Efficient GUI Agents: A Systems Survey of Observation, Memory, Action, and Runtime Optimization
  September 2, 2026 · arXiv
• STaR-KV: Spatio-Temporal Adaptive Re-weighting for KV Cache Compression in GUI Vision-Language Models
  June 1, 2026 · arXiv
• A11y-Compressor: A Framework for Enhancing the Efficiency of GUI Agent Observations through Visual Context Reconstruction and Redundancy Reduction
  May 1, 2026 · arXiv
• STEP: Success-Rate-Aware Trajectory-Efficient Policy Optimization
  November 17, 2025 · Findings of ACL 2026
• VRL-Bench: Benchmarking agents on computer control tasks under finite trial budgets
  September 11, 2026 · arXiv
• FinCUABuild: Can Agents Build Reliable Benchmarks for Dynamic Financial Computer Use?
  September 7, 2026 · arXiv
• Selective Knowledge Control for Continual GUI Agent Learning over Application Streams
  September 6, 2026 · arXiv
• AgentHijack: Visual Patch Attacks on Multimodal Computer-Use Agents
  September 6, 2026 · arXiv
• From Interaction Traces to Persistent Skills: Online Evolution for Computer-Use Agents
  September 4, 2026 · arXiv
• Do GUI Agents Know When Not to Act? Enabling Conflict-Aware Termination for Multimodal GUI Agents
  September 3, 2026 · arXiv
• SIR: Self-improving Red-teaming for Compute Use Agents
  August 31, 2026 · arXiv
• Iron: Intent-Aligned and Retrospective Dual Learning Framework for Enhancing Generalist Virtual Agents
  August 28, 2026 · arXiv
• UI-Venus-2 Technical Report
  August 27, 2026 · arXiv
• Task-Adaptive Rubrics for GUI Reward Modeling
  August 25, 2026 · arXiv
• CausalCache: Conditional High-Fidelity Restoration for Long-Horizon GUI Agents
  August 23, 2026 · arXiv
• GUI-Primitives: Diagnosing Spatial Reasoning Failures in Vision-Language GUI Grounding
  August 22, 2026 · arXiv
• Beyond Success and Failure: Length-Aware Contrastive Learning for GUI Agents
  August 22, 2026 · arXiv
• Software Engineering for and with GUI Agent
  August 10, 2026 · arXiv
• Hallucination-Free GUI Grounding via Regression-Free Layout-Aware Matching
  August 10, 2026 · arXiv
• GUI-Lens: Coarse-to-Fine Cropping for GUI Grounding with General-Purpose VLMs
  August 4, 2026 · arXiv
• Naive Visual Memory is Not Enough: A Failure-Mode Study of GUI Agents
  June 12, 2026 · arXiv
• Demo2Tutorial: From Human Experience to Multimodal Software Tutorials
  June 2, 2026 · CVPR 2026
• GUI-C²: Coarse-to-Fine GUI Grounding via Difficulty-Aware Reinforcement Learning
  May 29, 2026 · arXiv