Designing for all? Accessibility of native android interfaces from large language models

Exploring the Potential of Program Flowcharts on Code Generation Using Multimodal LLMs

In recent years, Large Language Models (LLMs) have made significant strides, leading to the emergence of multimodal LLMs capable of processing diverse inputs such as images and audio. Previous research indicates that the supply of multimodal LLMs with combined textual and visual information improves the automatic code generation capabilities. In software development, diagrams such as flowcharts are widely employed to facilitate tasks like code comprehension. While existing studies investigated the impact of visual inputs on LLMs and the usage of software diagrams, the potential influence of providing flowcharts on multimodal LLM performance remains underexplored. In this study, we generated flowcharts from example solution code for AtCoder problems and provided these visual aids alongside problem statements to GPT-4o for code generation. Our findings demonstrate that integrating flowcharts with problem statements yields performance improvements of up to 10%. Furthermore, when employing abstracted flowcharts, we observed a trend indicating that increasing levels of flowchart detail correlate with enhanced performance. Additionally, we compared the effectiveness of flowchart provision to Few-Shot Learning approaches. The findings suggest that one-shot learning provides sustainable improvements, whereas two-shot learning results in only minor improvements. Our work highlights the importance of software diagrams in supporting multimodal LLM-driven code generation.

A HYBRID APPROACH FOR REVERSE ENGINEERING GUI MODEL FROM MOBILE APPS FOR AUTOMATED TESTING

Unreliable in Practice? A Comprehensive Study of Errors in LLM-Generated Code

It is observed that generated code often omits basic input validation or memory-safety checks, which can lead to overflows, resource exhaustion, or other reliability/security issues, and even the largest models frequently make simple mistakes.

PROBE: Benchmarking code generation in large language models

The findings show that, while LLMs achieve promising results, they struggle with harder problems and with programming languages that have fewer available resources for training, and they often fail due to fundamental and easily avoidable errors that underscore the unreliability of automatically generated code.

From Design to Code: Exploring LLM-Supported Workflows in Figma Dev Mode

Design systems are widely used to ensure consistency and scalability in modern UI development, yet translating design artifacts into system-aligned code remains challenging. Developers must interpret visual designs and map them to appropriate components and implementation details, often relying on external documentation. While AI-assisted coding tools can accelerate development, they often lack structured design-system context, limiting their effectiveness in design-to-code workflows. In this paper, we investigate how a structured design context can influence developer experience. We present an empirical within-subject study with N=30 frontend developers comparing a traditional documentation-driven workflow to an MCP-supported workflow integrating Figma Dev Mode, Code Connect, and AI-assisted coding tools. The results show significant improvements in implementation accuracy, task efficiency, and perceived usability when a structured design context is available. Our findings suggest that AI-supported workflows can improve developer experience by reducing ambiguity in design interpretation while maintaining the need for human validation and documentation support.

An Empirical Study for Android-to-OpenHarmony GUI Test Migration

An enhanced approach based on ITeM is proposed, referred as ITeM-HM, which incorporates specific OpenHarmony system features and achieves a 214% success-rate relative improvement over the original ITeM, which is primarily hindered by OpenHarmony-specific characteristics.