OSWorld-Science: A Benchmark of Computer Use Agents for Learning and Using Scientific Software
cs.AI
Submitted: 2026-09-30
Updated: 2026-09-30
Code: https://github.com/DiscoAILab/OSWorld-Science
Terminology
Sources
- Windows Agent Arena: Evaluating Multi-Modal OS Agents at Scale
- OSWorld-MCP: Benchmarking MCP Tool Invocation In Computer-Use Agents
- MiniMax Sparse Attention
- PaperQA: Retrieval-Augmented Generative Agent for Scientific Research
- Benchmarking AI Agents for Addressing Scientific Challenges Across Scales
- Qwen-CUA: Native Computer Use for (almost) Everything
- AndroidWorld: A Dynamic Benchmarking Environment for Autonomous Agents
- ScienceBoard: Evaluating Multimodal Autonomous Agents in Realistic Scientific Workflows
- Kimi K3: Open Frontier Intelligence
- LogicVista: Multimodal LLM Logical Reasoning Benchmark in Visual Contexts
- OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments
- OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks
- WebArena: A Realistic Web Environment for Building Autonomous Agents
- The average of the first invariant factor for reductions of CM elliptic curves mod $p$
- A self-consistent Hartree-Fock approach for interacting bosons in optical lattices
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems