From the source

Introducing SWE-bench Verified

Research paper

OpenAI is releasing a human-validated subset of SWE-bench to better evaluate AI models on real-world software issues.

From the source
We’re releasing a human-validated subset of SWE-bench that more reliably evaluates AI models’ ability to solve real-world software issues.
openai.com

Other stories

OpenAI· 16 Oct

How Itaú is building generative AI for its entire customer base

Itaú's AI assistant and governance platform serve millions in a regulated sector

OpenAI· 16 OctCapability change

Stack Overflow for Agents triples signups with ChatGPT plugin

Plugin drives 3x registration surge and 10x agent creation

Announced: 13 Aug 2024 10:00 UTC
Detected: 03 Aug 2026 21:11 UTC (+720d after)

Visit source ↗openai.com

OpenAI· 15 Oct

Sullivan & Cromwell brings legal expertise into agreement review

Law firm builds AI tool for draft agreement review

OpenAI· 14 OctCapability change

SentinelOne cuts malware analysis from weeks to hours

SentinelOne uses OpenAI Daybreak to speed malware reverse engineering from weeks to hours.

Amp puts AI agents to work beyond code with OpenAI models
Amp uses OpenAI models for non-coding agent tasks
OpenAI· 14 Oct
How Block is building AI-powered security with OpenAI technology
Block uses OpenAI Daybreak to speed vulnerability fixes
OpenAI· 14 Oct