AIRS-Bench: a Suite of Tasks for Frontier AI Research Science Agents

• URL: http://arxiv.org/abs/2602.06855v2
• Date: Mon, 09 Feb 2026 14:39:19 GMT
• Title: AIRS-Bench: a Suite of Tasks for Frontier AI Research Science Agents
• Authors: Alisia Lupidi, Bhavul Gauri, Thomas Simon Foster, Bassel Al Omari, Despoina Magka, Alberto Pepe, Alexis Audran-Reiss, Muna Aghamelu, Nicolas Baldwin, Lucia Cipolina-Kun, Jean-Christophe Gagnon-Audet, Chee Hau Leow, Sandra Lefdal, Hossam Mossalam, Abhinav Moudgil, Saba Nazir, Emanuel Tewolde, Isabel Urrego, Jordi Armengol Estape, Amar Budhiraja, Gaurav Chaurasia, Abhishek Charnalia, Derek Dunfield, Karen Hambardzumyan, Daniel Izcovich, Martin Josifoski, Ishita Mediratta, Kelvin Niu, Parth Pathak, Michael Shvartsman, Edan Toledo, Anton Protopopov, Roberta Raileanu, Alexander Miller, Tatiana Shavrina, Jakob Foerster, Yoram Bachrach,
• Abstract summary: We introduce AIRS-Bench, a suite of 20 tasks sourced from state-of-the-art machine learning papers.<n>Airs-Bench tasks assess agentic capabilities over the full research lifecycle.<n>We open-source the AIRS-Bench task definitions and evaluation code to catalyze further development in autonomous scientific research.
• Score: 49.67355440164857
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: LLM agents hold significant promise for advancing scientific research. To accelerate this progress, we introduce AIRS-Bench (the AI Research Science Benchmark), a suite of 20 tasks sourced from state-of-the-art machine learning papers. These tasks span diverse domains, including language modeling, mathematics, bioinformatics, and time series forecasting. AIRS-Bench tasks assess agentic capabilities over the full research lifecycle -- including idea generation, experiment analysis and iterative refinement -- without providing baseline code. The AIRS-Bench task format is versatile, enabling easy integration of new tasks and rigorous comparison across different agentic frameworks. We establish baselines using frontier models paired with both sequential and parallel scaffolds. Our results show that agents exceed human SOTA in four tasks but fail to match it in sixteen others. Even when agents surpass human benchmarks, they do not reach the theoretical performance ceiling for the underlying tasks. These findings indicate that AIRS-Bench is far from saturated and offers substantial room for improvement. We open-source the AIRS-Bench task definitions and evaluation code to catalyze further development in autonomous scientific research.

Related papers

• Let the Barbarians In: How AI Can Accelerate Systems Performance Research [80.43506848683633]
  We term this iterative cycle of generation, evaluation, and refinement AI-Driven Research for Systems.<n>We demonstrate that ADRS-generated solutions can match or even outperform human state-of-the-art designs.
  arXiv  Detail & Related papers  (2025-12-16T18:51:23Z)
• InnovatorBench: Evaluating Agents' Ability to Conduct Innovative LLM Research [36.46396692622759]
  We introduce InnovatorBench, a benchmark-platform pair for realistic, end-to-end assessment of agents performing Large Language Model (LLM) research.<n>It comprises 20 tasks spanning Data Construction, Filtering, Augmentation, Loss Design, Reward Design, and Scaffold Construction.<n>We also develop ResearchGym, a research environment offering rich action spaces, distributed and long-horizon execution, asynchronous monitoring, and snapshot saving.
  arXiv  Detail & Related papers  (2025-10-31T16:22:23Z)
• AstaBench: Rigorous Benchmarking of AI Agents with a Scientific Research Suite [75.58737079136942]
  We present AstaBench, a suite that provides the first holistic measure of agentic ability to perform scientific research.<n>Our suite comes with the first scientific research environment with production-grade search tools.<n>Our evaluation of 57 agents across 22 agent classes reveals several interesting findings.
  arXiv  Detail & Related papers  (2025-10-24T17:10:26Z)
• RExBench: Can coding agents autonomously implement AI research extensions? [14.147417159347448]
  Agents based on Large Language Models (LLMs) have shown promise for performing sophisticated software engineering tasks autonomously.<n>We argue that research extension and its implementation is a critical capability for such systems.<n>We introduce RExBench to support the evaluation of this capability.
  arXiv  Detail & Related papers  (2025-06-27T19:41:41Z)
• R&D-Agent: An LLM-Agent Framework Towards Autonomous Data Science [70.1638335489284]
  High-level machine learning engineering tasks remain labor-intensive and iterative.<n>We introduce R&D-Agent, a comprehensive, decoupled, and framework that formalizes the machine learning process.<n>R&D-Agent defines the MLE into two phases and six components, turning agent design for MLE into a principled, testable process.
  arXiv  Detail & Related papers  (2025-05-20T06:07:00Z)
• MLGym: A New Framework and Benchmark for Advancing AI Research Agents [51.9387884953294]
  We introduce Meta MLGym and MLGym-Bench, a new framework and benchmark for evaluating and developing large language models on AI research tasks.<n>This is the first Gym environment for machine learning (ML) tasks, enabling research on reinforcement learning (RL) algorithms for training such agents.<n>We evaluate a number of frontier large language models (LLMs) on our benchmarks such as Claude-3.5-Sonnet, Llama-3.1 405B, GPT-4o, o1-preview, and Gemini-1.5 Pro.
  arXiv  Detail & Related papers  (2025-02-20T12:28:23Z)
• ML Research Benchmark [0.0]
  We present the ML Research Benchmark (MLRB), comprising 7 competition-level tasks derived from recent machine learning conference tracks. This paper introduces a novel benchmark and evaluates it using agent scaffolds powered by frontier models, including Claude-3 and GPT-4o. The results indicate that the Claude-3.5 Sonnet agent performs best across our benchmark, excelling in planning and developing machine learning models.
  arXiv  Detail & Related papers  (2024-10-29T21:38:42Z)
• CORE-Bench: Fostering the Credibility of Published Research Through a Computational Reproducibility Agent Benchmark [11.794931453828974]
  CORE-Bench is a benchmark consisting of 270 tasks based on 90 scientific papers across three disciplines (computer science, social science, and medicine) We provide an evaluation system to measure the accuracy of agents in a fast and parallelizable way. The best agent achieved an accuracy of 21% on the hardest task, showing the vast scope for improvement in automating routine scientific tasks.
  arXiv  Detail & Related papers  (2024-09-17T17:13:19Z)
• ComfyBench: Benchmarking LLM-based Agents in ComfyUI for Autonomously Designing Collaborative AI Systems [80.69865295743149]
  This work attempts to study using LLM-based agents to design collaborative AI systems autonomously.<n>Based on ComfyBench, we develop ComfyAgent, a framework that empowers agents to autonomously design collaborative AI systems by generating.<n>While ComfyAgent achieves a comparable resolve rate to o1-preview and significantly surpasses other agents on ComfyBench, ComfyAgent has resolved only 15% of creative tasks.
  arXiv  Detail & Related papers  (2024-09-02T17:44:10Z)
• MLAgentBench: Evaluating Language Agents on Machine Learning Experimentation [96.71370747681078]
  We introduce MLAgentBench, a suite of 13 tasks ranging from improving model performance on CIFAR-10 to recent research problems like BabyLM. For each task, an agent can perform actions like reading/writing files, executing code, and inspecting outputs. We benchmark agents based on Claude v1.0, Claude v2.1, Claude v3 Opus, GPT-4, GPT-4-turbo, Gemini-Pro, and Mixtral and find that a Claude v3 Opus agent is the best in terms of success rate.
  arXiv  Detail & Related papers  (2023-10-05T04:06:12Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.