跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.20145 · v1

Rank, Don't Regress: A 279M Text Ranker Beats 2B and 7B VLMs on Referential Urdu GUI Grounding

首次提交 2026-11-16 01:26
版本更新 2026-11-16 01:26
雷达观测 2026-11-17 12:06
arXiv ↗PDF ↗

摘要

UI grounding asks a model to find the screen element that a user instruction refers to. Most current systems use a vision-language model (VLM) that reads a screenshot and outputs pixel coordinates. We study this task on Urdu mobile web pages, a setting with no published UI grounding results. Instead of regressing coordinates, we take the candidate elements from the page DOM and train a 279M-parameter cross-encoder (mDeBERTa-v3-base) to pick one. The click point is the centre of the picked element's box, so it is exact whenever the right element is picked. We built a frozen test set of 304 instructions from six Urdu news sites; 149 of them (49%) are referential and never quote the full target text. Test domains were banned from training data before collection started. The ranker scores 73.4% point-in-box accuracy, against 29.6% for UI-TARS-2B and 23.7% for Qwen2.5-VL-7B, both zero-shot. On the referential slice the ranker scores 65.8%, against 30.2% for a string-matching baseline that needs no training. In a controlled comparison at small scale, with training data drawn from the same pages, a coordinate-regression model of similar size scores 1.6%. Adding element geometry as input tokens gives the best variant, at 75.7%. The ranker's errors concentrate in two places: elements that share identical text, and ordinal instructions such as "the fourth link".

作者备注:10 pages, 6 tables, 1 figure

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-16 01:26 · 最新观测版本

Rank, Don't Regress: A 279M Text Ranker Beats 2B and 7B VLMs on Referential Urdu GUI Grounding

UI grounding asks a model to find the screen element that a user instruction refers to. Most current systems use a vision-language model (VLM) that reads a screenshot and outputs pixel coordinates. We study this task on Urdu mobile web pages, a setting with no published UI grounding results. Instead of regressing coordinates, we take the candidate elements from the page DOM and train a 279M-parameter cross-encoder (mDeBERTa-v3-base) to pick one. The click point is the centre of the picked element's box, so it is exact whenever the right element is picked. We built a frozen test set of 304 instructions from six Urdu news sites; 149 of them (49%) are referential and never quote the full target text. Test domains were banned from training data before collection started. The ranker scores 73.4% point-in-box accuracy, against 29.6% for UI-TARS-2B and 23.7% for Qwen2.5-VL-7B, both zero-shot. On the referential slice the ranker scores 65.8%, against 30.2% for a string-matching baseline that needs no training. In a controlled comparison at small scale, with training data drawn from the same pages, a coordinate-regression model of similar size scores 1.6%. Adding element geometry as input tokens gives the best variant, at 75.7%. The ranker's errors concentrate in two places: elements that share identical text, and ordinal instructions such as "the fourth link".

查看此版本解读 →查看此版本原文 ↗