AI Paper Atlas

Research map / Large language models / Agents, code and math

FeatureBench: Benchmarking Agentic Coding for Complex Feature Development

Qixing Zhou, JiaCheng Zhang, Haiyang Wang, Rui Hao, Jiahe Wang, Minghao Han, Yuxue Yang, Shuzhe Wu, Feiyang Pan, Lue Fan, Dandan Tu, Zhaoxiang Zhang

ICLR 2026 · Large language models › Agents, code and math · approach: Software engineering agents · 27 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Add interface definitions to the task (ablation)Gemini-3-Pro-Preview (OpenHands) · FeatureBench Lite% resolved: 3.3 -> 10.0Improved
Add interface definitions to the task (ablation)GPT-5.1-Codex · FeatureBench Lite% resolved: 16.7 -> 20.0Improved
Make unit tests visible to the agent (ablation)Gemini-3-Pro-Preview (OpenHands) · FeatureBench Lite% resolved: 10.0 -> 60.0Improved
Make unit tests visible to the agent (ablation)GPT-5.1-Codex · FeatureBench Lite% resolved: 20.0 -> 63.3Improved
Evaluate on FeatureBench instead of SWE-bench VerifiedClaude Opus 4.5 · resolved rate% resolved: 74.40 -> 5.2Hurt
Use Claude Opus 4.5 instead of Qwen3-Coder under OpenHandsOpenHands · FeatureBench Full% resolved: 3.5 -> 10.5Improved
Use Lite subset instead of Full setall configurations · FeatureBenchrankingNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy