# Breakpoint: Scalable evaluation of system-level reasoning in LLM code agents

arXiv: https://arxiv.org/abs/2506.00172 · HTML: https://mlsanity.com/paper/2506.00172/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning.
- Builds on 11 of its 14 references that are in the map.
- First posted on arXiv 30 May 2025.

## Placed in

- LLM reasoning / [Reasoning for agents, planning and code](https://mlsanity.com/lr/reasoning-for-agents-planning-and-code/)
- LLM agents and tool use / [Software engineering coding agents](https://mlsanity.com/agents/software-engineering-coding-agents/)

## Builds on (most cited first, 11 of 11)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering](https://mlsanity.com/paper/2405.15793/) (2024)
- [Solving Quantitative Reasoning Problems with Language Models](https://mlsanity.com/paper/2206.14858/) (2022)
- [AgentBench: Evaluating LLMs as Agents](https://mlsanity.com/paper/2308.03688/) (2023)
- [FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AI](https://mlsanity.com/paper/2411.04872/) (2024)
- [Measuring AI Ability to Complete Long Software Tasks](https://mlsanity.com/paper/2503.14499/) (2025)
- [LLM Critics Help Catch LLM Bugs](https://mlsanity.com/paper/2407.00215/) (2024)
- [SWE-Lancer: Can Frontier LLMs Earn 1 Million from Real-World Freelance Software Engineering?](https://mlsanity.com/paper/2502.12115/) (2025)
- [Towards Long-Horizon Vision-Language Navigation: Platform, Benchmark and Method](https://mlsanity.com/paper/2412.09082/) (2024)
- [MARPLE: A Benchmark for Long-Horizon Inference](https://mlsanity.com/paper/2410.01926/) (2024)

## Built on by (most cited first, 5 of 5)

- [The Landscape of Agentic Reinforcement Learning for LLMs: A Survey](https://mlsanity.com/paper/2509.02547/) (2025)
- [ImpossibleBench: Measuring LLMs' Propensity of Exploiting Test Cases](https://mlsanity.com/paper/2510.20270/) (2025)
- [GeoChallenge: A Multi-Answer Multiple-Choice Benchmark for Geometric Reasoning with Diagrams](https://mlsanity.com/paper/2603.19252/) (2026)
- [Evaluating RL Explainability Methods by How Much They Help Fix Bugs in Agents](https://mlsanity.com/paper/2608.17524/) (2026)
- [Tev: Scaling Long-Horizon Agentic Environments with Dense Progress Rewards](https://mlsanity.com/paper/2610.18868/) (2026)

## Abstract (arXiv)

Benchmarks for large language models (LLMs) have predominantly assessed short-horizon, localized reasoning. Existing long-horizon suites (e.g. SWE-bench) rely on manually curated issues, so expanding or tuning difficulty demands expensive human effort and evaluations quickly saturate. However, many real-world tasks, such as software engineering or scientific research, require agents to rapidly comprehend and manipulate novel, complex structures dynamically; evaluating these capabilities requires the ability to construct large and varied sets of problems for agents to solve. We introduce Breakpoint, a benchmarking methodology that automatically generates code-repair tasks by adversarially corrupting functions within real-world software repositories. Breakpoint systematically controls task difficulty along two clear dimensions: local reasoning (characterized by code complexity metrics such as cyclomatic complexity) and system-level reasoning (characterized by call-graph centrality and the number of simultaneously corrupted interdependent functions). In experiments across more than 900 generated tasks we demonstrate that our methodology can scale to arbitrary difficulty, with state-of-the-art models' success rates ranging from 55% on the easiest tasks down to 0% on the hardest.