TRACE: Trajectory-robust Admission with Evidence Ordering for Efficient GUI Agents

Hugging Face Daily Papers Papers
gui-agents visual-token-pruning efficiency training-free token-ordering coverage-aware kv-contraction

Summary

TRACE is a training-free framework that optimizes GUI agent efficiency by ranking visual evidence based on utility and diversity, reducing latency and memory usage through adaptive token management and KV contraction.

GUI agents accumulate high-resolution screenshots as the trajectory unfolds, increasing inference latency and memory usage. Training-free visual token pruning can reduce this cost, but cache reuse introduces a fundamental constraint. Once tokens are discarded, the corresponding visual evidence cannot be recovered without re-encoding. Pruning therefore becomes an irreversible admission decision that must remain useful for unknown future targets while preserving coverage of operable regions under tight budgets. To address these challenges, we propose \method{}, a training-free framework for \textbf{Trajectory-robust Admission and Coverage-aware Evidence ordering}. Specifically, we combine a query-independent layout-derived interaction prior with instruction relevance and feature novelty to rank visual evidence according to both potential future utility and diversity. Then, we reserve part of the budget for native visual tokens distributed across the screen, repairing missing spatial coverage without breaking the ordering. Together, these mechanisms produce a nested token order, allowing retained visual evidence to shrink monotonically across budgets while remaining reusable throughout the trajectory. Finally, our monotone KV contraction incrementally contracts retired frames into compact session state, avoiding repeated visual encoding or pruning. Extensive experiments across six GUI benchmarks and diverse models verify the effectiveness of our proposed under tight budgets. The source code will be released.
Original Article
View Cached Full Text

Cached at: 09/14/26, 02:35 PM

Paper page - TRACE: Trajectory-robust Admission with Evidence Ordering for Efficient GUI Agents

Source: https://huggingface.co/papers/2609.10297

Abstract

TRACE is a training-free framework that ranks visual evidence by future utility and diversity, reserves native tokens for spatial coverage, and contracts retired frames to reduce latency and memory in GUI agents.

GUI agentsaccumulate high-resolution screenshots as the trajectory unfolds, increasing inference latency and memory usage. Training-freevisual token pruningcan reduce this cost, but cache reuse introduces a fundamental constraint. Once tokens are discarded, the corresponding visual evidence cannot be recovered without re-encoding. Pruning therefore becomes an irreversible admission decision that must remain useful for unknown future targets while preserving coverage of operable regions under tight budgets. To address these challenges, we propose \method{}, a training-free framework for \textbf{Trajectory-robust AdmissionandCoverage-aware Evidence ordering}. Specifically, we combine a query-independent layout-derivedinteraction priorwith instruction relevance and feature novelty to rank visual evidence according to both potential future utility and diversity. Then, we reserve part of the budget for native visual tokens distributed across the screen, repairing missing spatial coverage without breaking the ordering. Together, these mechanisms produce anested token order, allowing retained visual evidence to shrink monotonically across budgets while remaining reusable throughout the trajectory. Finally, our monotoneKV contractionincrementally contracts retired frames into compact session state, avoiding repeated visual encoding or pruning. Extensive experiments across six GUI benchmarks and diverse models verify the effectiveness of our proposed under tight budgets. The source code will be released.

View arXiv pageView PDFGitHubAdd to collection

Get this paper in your agent:

hf papers read 2609\.10297

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2609.10297 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.10297 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.10297 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles