---
title: "SWE-Bench ProMax: Benchmarking Agents on Large-Scale Multilingual Code Refactoring"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2608.09802v1"
permalink: "https://feed.astralor.com/items/itm_bed8a5a9110e1e2429bccb18"
published: "2026-08-10T16:23:19.000Z"
score: 77
channel: "ai"
tags: ["AI coding agent","Hugging Face","SWE-Bench ProMax","Code refactoring"]
---

# SWE-Bench ProMax: Benchmarking Agents on Large-Scale Multilingual Code Refactoring

## AI commentary

> The saturation and flaws of existing benchmarks are no secret, and targeting cross-file refactoring—something rarely stressed seriously—is a much more realistic direction. For researchers and agent developers, this is a valuable new yardstick; for ordinary users, it is a reminder that reported resolve rates may overestimate real engineering capability.

## At a glance

Existing AI coding agent benchmarks are rapidly saturating, and their evaluation quality has come under serious scrutiny: an audit found that nearly 60% of unsolved SWE-bench Verified instances contain flawed tests, and frontier models can reproduce gold patches verbatim from training data. To address this, researchers introduce SWE-Bench ProMax, an expert-curated multilingual code refactoring benchmark with 170 instances drawn from real commits across seven programming languages. On average, each instance involves modifying 11.4 files and 261.6 lines of code. Experiments with frontier models under two agent scaffolds show that the best model achieves only a 41.2% resolve rate, confirming that the benchmark offers a meaningful and unsaturated challenge for current AI coding agents. The dataset is publicly available on Hugging Face.

## Key points

- Audits reveal flawed tests and training-data leakage in existing benchmarks, raising concerns about evaluation quality
- SWE-Bench ProMax contains 170 real code refactoring tasks across seven languages, averaging 11.4 changed files and 261.6 lines per instance
- All instances undergo expert multi-stage curation, with rewritten problem descriptions and manually reviewed test suites
- The best frontier model achieves only 41.2% resolve rate under two agent scaffolds, showing the benchmark is not yet saturated

## Full text

As AI coding agents take on increasingly complex, long-horizon software engineering tasks, existing benchmarks are rapidly saturating and their evaluation quality has come under serious scrutiny: a recent audit found that nearly 60% of unsolved SWE-bench Verified instances contain flawed tests -- either overly narrow tests that reject correct solutions or overly broad tests that check unstated requirements -- and that frontier models can verbatim reproduce gold patches from training data. Code refactoring, which requires coordinated, behavior-preserving changes across many files, offers a substantially harder and more realistic test of agent capability, yet remains underserved by current benchmarks. We introduce SWE-Bench ProMax, an expert-curated, multilingual code refactoring benchmark of 170 instances drawn from real commits across seven programming languages (Python, Java, TypeScript, Go, C, C++, and Rust). Every instance undergoes rigorous, multi-stage curation that directly addresses the quality problems identified in prior benchmarks: issue descriptions are rewritten from scratch to provide precise, unambiguous specifications, and test suites are manually reviewed to remove overly narrow and overly broad tests. Tasks with insufficient complexity or limited cross-file scope are filtered out, yielding a benchmark of challenging, large-scale refactoring tasks that average 11.4 modified files and 261.6 lines of code per instance, substantially exceeding the scale of existing benchmarks. Experiments with frontier models under two agent scaffolds show that the best model achieves only 41.2% resolve rate, confirming that SWE-Bench ProMax presents a meaningful and unsaturated challenge for current AI coding agents. Our benchmark is available at https://huggingface.co/datasets/swe-bench-promax/SWE-Bench-ProMax.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_bed8a5a9110e1e2429bccb18