PulseAugur
EN
LIVE 09:04:25

New GUI Grounding Method Learns from Negative Samples

By PulseAugur Editorial · [1 sources] ·

Researchers have developed a new label-free test-time training method for GUI grounding, a crucial step for autonomous agents to interpret natural language commands. The approach, called Confidence-Anchored Negative Learning (CANL), leverages confidence patterns in coordinate tokens and prioritizes learning from negative samples over potentially noisy positive ones. CANL-7B demonstrated significant improvements, achieving 92.1% on the ScreenSpot-V2 benchmark and a 8.9% absolute improvement on the more challenging ScreenSpot-Pro dataset. AI

IMPACT This new method could significantly reduce the annotation costs for training autonomous agents, enabling more scalable development of GUI-grounding capabilities.

RANK_REASON Academic paper detailing a new method for GUI grounding. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

New GUI Grounding Method Learns from Negative Samples

How we ranked this

Signal score
14 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
Academic paper detailing a new method for GUI grounding. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [1]

  1. arXiv cs.CV TIER_1 English(EN) · Yizhou Liu, Fei Tang, Yuchen Yan, Zhengxi Lu, Songqin Nong, Tao Jiang, Wenhao Xu, Wenqi Zhang, Weiming Lu, Jun Xiao, Yongliang Shen ·

    Learning from Reliable Negatives: Confidence-Anchored Test-Time Adaptation for GUI Grounding

    arXiv:2609.15307v1 Announce Type: new Abstract: Graphical User Interface (GUI) grounding is essential for autonomous agents to map natural language instructions to precise screen coordinates. However, existing supervised fine-tuning and reinforcement learning methods are constrai…

RELATED ENTITIES

RELATED TOPICS