List पर वापस जाएँ
📄Papers

Learning to Detect UI Principle Violations via Reinforcement Learning

Mehta N. et al.
2026-07

परिचय

जुलाई 2026 का paper, जो generated web interfaces के लिए lightweight vision-language critic पर केंद्रित है। Authors WCAG 2.2 accessibility, deceptive design और cognitive usability को समेटने वाले 19 principles को एक साथ लाते हैं, फिर लगभग 10,000 generated pages पर verified synthetic violations के साथ training करते हैं।

सारांश

यह paper क्या नया संकेत देता है:

1. Correctness से आगे quality: केवल code चलने की बजाय जाँचता है कि generated interfaces उपयोगी और जिम्मेदार भी हैं या नहीं
2. व्यापक design coverage: Accessibility, deceptive-pattern, perception, cognition और interaction principles को जोड़ता है
3. Reproducible data pipeline: अलग-अलग LLM-generated Tailwind pages में ज्ञात violations डालकर verify करता है
4. Efficient critic: 19-principle benchmark पर 4B vision-language model को 36% से 84% micro-F1 तक सुधारता है
5. Generation-loop utility: बेहतर UI generators के लिए critic को audit gate, training-data filter और reward signal की तरह रखता है

Tags

ui-evaluationaccessibilitydesign-qualityvision-language-modelreinforcement-learning
मूल लेख पढ़ें

मिलते-जुलते लेख

📄Papers

ControlAnything: Generative UI and On-Demand Device Integration in Mixed Reality

November 2026 preprint generating both spatial controls and the connector code for physical devices when a Quest 3 wearer requests them. A coding agent tests connections over hand-written MIDI, Bluetooth, and key-press transports, then returns AR widgets, a per-frame hand program, or both.

generative-uimixed-realitydevice-integration
लेखक Bektur Ryskeldiev · 2026-11-05
और पढ़ें
📄Papers

Interface Quilt: Modifying Applications with Computer Use Agents and Generative UI

November 2026 preprint from Autodesk Research combining computer-use agents and GenUI to build persistent interface patches over existing applications. It recovers controls from rendered pixels, generates new interaction surfaces, and synchronizes them with the underlying app without requiring source code, APIs, or accessibility data.

generative-uicomputer-use-agentsinterface-patches
लेखक Qiu L. et al. · 2026-11-01
और पढ़ें
📄Papers

ABRA: Agent-Bound UI for Human-Agent Communication beyond Chat with Reusable Views

October 2026 EMNLP system demonstration separating reusable graph, table, text, and display views from task-specific interaction meanings. ABRA publishes these bindings as shared canvas state so users and other agents can inspect and invoke task-specific actions.

generative-uishared-canvasesinteraction-semantics
लेखक Kim S. G. and Kim H. · 2026-10-22
और पढ़ें

Generative UI और Intelligent UI resources का curated संग्रह।

बनाया गया ❤️ community द्वारा