CODA-BENCH Evaluates Code Agents
jiqizhixin · x · 2026-07-12
A research team from Renmin University of China released CODA-BENCH, designed to simultaneously test coding capabilities and data intelligence.
- Task scenarios require agents to navigate complex file systems, hundreds of Kaggle datasets, and complete real-world data analysis coding tasks.
- Results show that even top-tier systems only achieved a success rate of 61.1% across 1,009 tasks, indicating that current agents still have significant shortcomings in data-intensive work.
- The post includes links to the project, paper, code, and dataset, along with a report from Jiqizhixin.
More from coding & agent
- GPT-6 Astra beats Factorio with enemies in 44 in-game hours at ~$4,500 API cost — liminal_bardo · 2026-09-11
- Investment Analyst Asks How to Build a Claude-Based Diligence Agent Stack — Careless_Tie2286 · 2026-09-11
- How Do You Catch Behavioral Regressions in LLM Agents Between Releases? — Beautiful_Belt_601 · 2026-09-11
- Treating agents like 50 First Dates: a 3-layer context system so every conversation doesn't start from zero — evielync · 2026-09-11
- Running the Firefox MCP on Android via Termux, ngrok, and mcp-proxy — Nervous-Strain7544 · 2026-09-11
- Run Firefox MCP on Android: Termux + ngrok tunnel tutorial — Nervous-Strain7544 · 2026-09-11