# Asuka-Bench: Benchmarking Code Agents on Underspecified User Intent and Multi-Round Refinement

arXiv: https://arxiv.org/abs/2606.05920 · HTML: https://mlsanity.com/paper/2606.05920/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use.
- Builds on 12 of its 24 references that are in the map.
- First posted on arXiv 4 Jun 2026.

## Placed in

- LLM agents and tool use / [Software engineering coding agents](https://mlsanity.com/agents/software-engineering-coding-agents/)

## Builds on (most cited first, 12 of 12)

- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools](https://mlsanity.com/paper/2406.12793/) (2024)
- [Design2Code: Benchmarking Multimodal Code Generation for Automated Front-End Engineering](https://mlsanity.com/paper/2403.03163/) (2024)
- [WebGen-Bench: Evaluating LLMs on Generating Interactive and Functional Websites from Scratch](https://mlsanity.com/paper/2505.03733/) (2025)
- [ArtifactsBench: Bridging the Visual-Interactive Gap in LLM Code Generation Evaluation](https://mlsanity.com/paper/2507.04952/) (2025)
- [Web-Bench: A LLM Code Benchmark Based on Web Standards and Frameworks](https://mlsanity.com/paper/2505.07473/) (2025)
- [WebMMU: A Benchmark for Multimodal Multilingual Website Understanding and Code Generation](https://mlsanity.com/paper/2508.16763/) (2025)
- [WebUIBench: A Comprehensive Benchmark for Evaluating Multimodal Large Language Models in WebUI-to-Code](https://mlsanity.com/paper/2506.07818/) (2025)
- [Automatically Benchmarking LLM Code Agents through Agent-Driven Annotation and Evaluation](https://mlsanity.com/paper/2510.24358/) (2025)

## Built on by (most cited first, 4 of 4)

- [SWE-INTERACT: Reimagining SWE Benchmarks as User-Driven Long-Horizon Coding Sessions](https://mlsanity.com/paper/2606.30573/) (2026)
- [What Do We Expect from LLMs? Mapping the Design of LLM Benchmarks](https://mlsanity.com/paper/2609.19182/) (2026)
- [EvoWeb: A Benchmark for Coding Agents on Evolving Web Applications](https://mlsanity.com/paper/2610.16402/) (2026)
- [Instructions Are Evidence, Not Objectives: Measuring Outcome Alignment of Agents](https://mlsanity.com/paper/2610.17757/) (2026)

## Abstract (arXiv)

Existing code-generation benchmarks score a single mapping from a complete prompt to a one-shot output. However, real web development is different. Users seldom write a full spec at the start; many requirements only become clear once they look at an intermediate result and react to it. We present Asuka-Bench, a benchmark that pairs underspecified user intent with multi-round refinement, grounded in browser-rendered behavior. Each task is resolved through a closed loop: a Code Agent generates a web project, a UI Agent executes test cases on the deployed site, and a User LLM turns evaluation outcomes into natural-language feedback for the next round. The benchmark comprises 50 web tasks with 784 evaluation criteria and 2402 expected outcomes. We benchmark 8 LLMs across 2 agent frameworks. The results separate models clearly: weighted Task Pass Rate varies by 38 percentage points and models also differ substantially in their ability to repair from feedback. Asuka-Bench is also far from saturated: even the strongest model completes only 52% of projects after three rounds.