Can a coding agent take over a teammate’s work in progress?
Introducing SWE-CoWork, a benchmark for coding agents working as digital employees on long-horizon software engineering tasks.
Each assignment starts with a short handoff in a simulated workplace modeled on real-world engineering workflows. Agents search task boards and versioned docs, ask teammates for missing details, and revise their code when a reply changes what needs to be built.
Open questions and confirmed requirements can both be lost before the agent hands the work back. SWE-CoWork lets us trace them through the conversation, the patch, and the tests.
101 assignments across 101 real open-source repositories.
26 simulated teammates per workplace.
11 programming languages.
12 application domains.
17 workplace tools.
🌐 Website: https://t.co/vyddiAIGDE
📄 Paper: https://t.co/EPBQ5RjlGy
💻 Code: https://t.co/zU8wPB0Wir
🤗 Hugging Face Daily Papers: https://t.co/tAbAQZ4pKR (If you find it useful, we’d appreciate an upvote!)
[1/9]
What are the limits of AI agents in scientific research? Can they use professional software as effectively as scientists?
Meet OSWorld-Science: a benchmark and evaluation environment for computer-use agents, with tasks grounded in real-world scientific research needs—from pathology image analysis to physical simulation.
146 tasks.
7 scientific domains.
17 tools.
12 VLMs evaluated.
Tasks are developed through expert proposals and iterative human–AI co-design, with selection guided by scientific value and difficulty.
By combining GUI/CLI interaction, artifact-based evaluation, and trajectory analysis, we examine how agents carry out scientific workflows and where execution breaks down.
TokenWave is glad to be part of this research collaboration.
🌐 Project: https://t.co/XK0B87AGzB
📄 Paper: https://t.co/c4RHtcKPxn
💻 Code: https://t.co/HW28bONUTr
📊 Data: https://t.co/95y8wzLghC
0/n
# Toward Human-Like RSI
Blog / project: https://t.co/FjwP5Pt6ZL
Recursive self-improvement (RSI) is often pictured as a model repeatedly rewriting its own weights against a fixed objective. Humans improve differently: we decide what “better” means when the goal is vague, test ourselves, judge what happened, and change both what we know and the tools we work with.
This three-paper series turns those parts of improvement into controlled, auditable research questions:
ASPIRE studies how agents decide *what* to learn from a vague goal;
S³Gym studies how they test, judge, and learn from noisy self-generated experience;
HarnessDev studies how they build and evolve the external systems that shape their behavior.
These benchmarks use hidden evaluation as a proxy for whether a change transfers beyond the agent's own local signals - they do not claim to demonstrate open-ended RSI.