Kilian Lieret

Talor Abramovich, Meet Udeshi, Minghao Shao et al. · princeton, uw

Although language model (LM) agents have demonstrated increased performance in multiple domains, including coding and web-browsing, their success in cybersecurity has been limited. We present EnIGMA, an LM agent for autonomously solving Capture The Flag (CTF) challenges. We introduce new tools and interfaces to improve the agent's ability to find and exploit security vulnerabilities, focusing on interactive terminal programs. These novel Interactive Agent Tools enable LM agents, for the first time, to run interactive utilities, such as a debugger and a server connection tool, which are essential for solving these challenges. Empirical analysis on 390 CTF challenges across four benchmarks demonstrate that these new tools and interfaces substantially improve our agent's performance, achieving state-of-the-art results on NYU CTF, Intercode-CTF, and CyBench. Finally, we analyze data leakage, developing new methods to quantify it and identifying a new phenomenon we term soliloquizing, where the model self-generates hallucinated observations without interacting with the environment. Our code and development dataset are available at https://github.com/SWE-agent/SWE-agent/tree/v0.7 and https://github.com/NYU-LLM-CTF/NYU_CTF_Bench/tree/main/development respectively.

Minyang Tian, Luyu Gao, Shizhuo Dylan Zhang et al. · princeton, uw

Since language models (LMs) now outperform average humans on many challenging tasks, it has become increasingly difficult to develop challenging, high-quality, and realistic evaluations. We address this issue by examining LMs' capabilities to generate code for solving real scientific research problems. Incorporating input from scientists and AI researchers in 16 diverse natural science sub-fields, including mathematics, physics, chemistry, biology, and materials science, we created a scientist-curated coding benchmark, SciCode. The problems in SciCode naturally factorize into multiple subproblems, each involving knowledge recall, reasoning, and code synthesis. In total, SciCode contains 338 subproblems decomposed from 80 challenging main problems. It offers optional descriptions specifying useful scientific background information and scientist-annotated gold-standard solutions and test cases for evaluation. Claude3.5-Sonnet, the best-performing model among those tested, can solve only 4.6% of the problems in the most realistic setting. We believe that SciCode demonstrates both contemporary LMs' progress towards becoming helpful scientific assistants and sheds light on the development and evaluation of scientific AI in the future.

John Yang, Kilian Lieret, Joyce Yang et al.

Current benchmarks for coding evaluate language models (LMs) on concrete, well-specified tasks such as fixing specific bugs or writing targeted tests. However, human programmers do not spend all day incessantly addressing isolated tasks. Instead, real-world software development is grounded in the pursuit of high-level goals, like improving user retention or reducing costs. Evaluating whether LMs can also iteratively develop code to better accomplish open-ended objectives without any explicit guidance remains an open challenge. To address this, we introduce CodeClash, a benchmark where LMs compete in multi-round tournaments to build the best codebase for achieving a competitive objective. Each round proceeds in two phases: agents edit their code, then their codebases compete head-to-head in a code arena that determines winners based on objectives like score maximization, resource acquisition, or survival. Whether it's writing notes, scrutinizing documentation, analyzing competition logs, or creating test suites, models must decide for themselves how to improve their codebases both absolutely and against their opponents. We run 1680 tournaments (25,200 rounds total) to evaluate 8 LMs across 6 arenas. Our results reveal that while models exhibit diverse development styles, they share fundamental limitations in strategic reasoning. Models also struggle with long-term codebase maintenance, as repositories become progressively messy and redundant. These limitations are stark: top models lose every round against expert human programmers. We open-source CodeClash to advance the study of autonomous, goal-oriented code development.

John Yang, Kilian Lieret, Carlos E. Jimenez et al. · gatech, princeton

Despite recent progress in Language Models (LMs) for software engineering, collecting training data remains a significant pain point. Existing datasets are small, with at most 1,000s of training instances from 11 or fewer GitHub repositories. The procedures to curate such datasets are often complex, necessitating hundreds of hours of human labor; companion execution environments also take up several terabytes of storage, severely limiting their scalability and usability. To address this pain point, we introduce SWE-smith, a novel pipeline for generating software engineering training data at scale. Given any Python codebase, SWE-smith constructs a corresponding execution environment, then automatically synthesizes 100s to 1,000s of task instances that break existing test(s) in the codebase. Using SWE-smith, we create a dataset of 50k instances sourced from 128 GitHub repositories, an order of magnitude larger than all previous works. We train SWE-agent-LM-32B, achieving 40.2% Pass@1 resolve rate on the SWE-bench Verified benchmark, state of the art among open source models. We open source SWE-smith (collection procedure, task instances, trajectories, models) to lower the barrier of entry for research in LM systems for automated software engineering. All assets available at https://swesmith.com.

Ori Press, Brandon Amos, Haoyu Zhao et al. · princeton, uw

Despite progress in language model (LM) capabilities, evaluations have thus far focused on models' performance on tasks that humans have previously solved, including in programming (Jimenez et al., 2024) and mathematics (Glazer et al., 2024). We therefore propose testing models' ability to design and implement algorithms in an open-ended benchmark: We task LMs with writing code that efficiently solves computationally challenging problems in computer science, physics, and mathematics. Our AlgoTune benchmark consists of 154 coding tasks collected from domain experts and a framework for validating and timing LM-synthesized solution code, which is compared to reference implementations from popular open-source packages. In addition, we develop a baseline LM agent, AlgoTuner, and evaluate its performance across a suite of frontier models. AlgoTuner uses a simple, budgeted loop that edits code, compiles and runs it, profiles performance, verifies correctness on tests, and selects the fastest valid version. AlgoTuner achieves an average 1.72x speedup against our reference solvers, which use libraries such as SciPy, sk-learn and CVXPY. However, we find that current models fail to discover algorithmic innovations, instead preferring surface-level optimizations. We hope that AlgoTune catalyzes the development of LM agents exhibiting creative problem solving beyond state-of-the-art human performance.

John Yang, Kilian Lieret, Jeffrey Ma et al.

Turning ideas into full software projects from scratch has become a popular use case for language models. Agents are being deployed to seed, maintain, and grow codebases over extended periods with minimal human oversight. Such settings require models to make high-level software architecture decisions. However, existing benchmarks measure focused, limited tasks such as fixing a single bug or developing a single, specified feature. We therefore introduce ProgramBench to measure the ability of software engineering agents to develop software holisitically. In ProgramBench, given only a program and its documentation, agents must architect and implement a codebase that matches the reference executable's behavior. End-to-end behavioral tests are generated via agent-driven fuzzing, enabling evaluation without prescribing implementation structure. Our 200 tasks range from compact CLI tools to widely used software such as FFmpeg, SQLite, and the PHP interpreter. We evaluate 9 LMs and find that none fully resolve any task, with the best model passing 95\% of tests on only 3\% of tasks. Models favor monolithic, single-file implementations that diverge sharply from human-written code.

Kilian Lieret, Gage DeZoort, Devdoot Chatterjee et al. · princeton

Recent work has demonstrated that graph neural networks (GNNs) can match the performance of traditional algorithms for charged particle tracking while improving scalability to meet the computing challenges posed by the HL-LHC. Most GNN tracking algorithms are based on edge classification and identify tracks as connected components from an initial graph containing spurious connections. In this talk, we consider an alternative based on object condensation (OC), a multi-objective learning framework designed to cluster points (hits) belonging to an arbitrary number of objects (tracks) and regress the properties of each object. Building on our previous results, we present a streamlined model and show progress toward a one-shot OC tracking algorithm in a high-pileup environment.

Shitij Govil, Jack P. Rodgers, Yuan-Tang Chou et al.

Charged particle track reconstruction is a foundational task in collider experiments and the main computational bottleneck in particle reconstruction. Graph neural networks (GNNs) have shown strong performance for this problem, but costly graph construction, irregular computations, and random memory access patterns substantially limit their throughput. The recently proposed Hashing-based Efficient Point Transformer (HEPT) offers a theoretically guaranteed near-linear complexity for large point cloud processing via locality-sensitive hashing (LSH) in attention computations; however, its evaluations have largely focused on embedding quality, and the object condensation pipeline on which HEPT relies requires a post-hoc clustering step (e.g., DBScan) that can dominate runtime. In this work, we make two contributions. First, we present a unified, fair evaluation of physics tracking performance for HEPT and a representative GNN-based pipeline under the same dataset and metrics. Second, we introduce HEPTv2 by extending HEPT with a lightweight decoder that eliminates the clustering stage and directly predicts track assignments. This modification preserves HEPT's regular, hardware-friendly computations while enabling ultra-fast end-to-end inference. On the TrackML dataset, optimized HEPTv2 achieves approximately 28 ms per event on an A100 while maintaining competitive tracking efficiency. These results position HEPTv2 as a practical, scalable alternative to GNN-based pipelines for fast tracking.

John Yang, Carlos E. Jimenez, Alexander Wettig et al.

Language model (LM) agents are increasingly being used to automate complicated tasks in digital environments. Just as humans benefit from powerful software applications, such as integrated development environments, for complex tasks like software engineering, we posit that LM agents represent a new category of end users with their own needs and abilities, and would benefit from specially-built interfaces to the software they use. We investigate how interface design affects the performance of language model agents. As a result of this exploration, we introduce SWE-agent: a system that facilitates LM agents to autonomously use computers to solve software engineering tasks. SWE-agent's custom agent-computer interface (ACI) significantly enhances an agent's ability to create and edit code files, navigate entire repositories, and execute tests and other programs. We evaluate SWE-agent on SWE-bench and HumanEvalFix, achieving state-of-the-art performance on both with a pass@1 rate of 12.5% and 87.7%, respectively, far exceeding the previous state-of-the-art achieved with non-interactive LMs. Finally, we provide insight on how the design of the ACI can impact agents' behavior and performance.

Daniel S. Katz, Sudhir Malik, Mark S. Neubauer et al.

New facilities of the 2020s, such as the High Luminosity Large Hadron Collider (HL-LHC), will be relevant through at least the 2030s. This means that their software efforts and those that are used to analyze their data need to consider sustainability to enable their adaptability to new challenges, longevity, and efficiency, over at least this period. This will help ensure that this software will be easier to develop and maintain, that it remains available in the future on new platforms, that it meets new needs, and that it is as reusable as possible. This report discusses a virtual half-day workshop on "Software Sustainability and High Energy Physics" that aimed 1) to bring together experts from HEP as well as those from outside to share their experiences and practices, and 2) to articulate a vision that helps the Institute for Research and Innovation in Software for High Energy Physics (IRIS-HEP) to create a work plan to implement elements of software sustainability. Software sustainability practices could lead to new collaborations, including elements of HEP software being directly used outside the field, and, as has happened more frequently in recent years, to HEP developers contributing to software developed outside the field rather than reinventing it. A focus on and skills related to sustainable software will give HEP software developers an important skill that is essential to careers in the realm of software, inside or outside HEP. The report closes with recommendations to improve software sustainability in HEP, aimed at the HEP community via IRIS-HEP and the HEP Software Foundation (HSF).