cd/entity/Frontier Evals· home› entities› Frontier Evals
grep -l @frontier evals /news/*.json | wc -l → 1

Frontier Evals

mentions 1 type Person feed RSS

// recent coverage 1 mentions

11:10
2026-06-24
byteiota.com
ai-research

SWE-bench Pro: How to Read the Coding Agent Leaderboard

OpenAI abandoned SWE-bench Verified on February 23, 2026, after finding 59.4% of its hardest failed tests were broken and training data contamination inflated scores. Its replacement, SWE-bench Pro fr…

// co-occurs with top 7 entities
OpenAI 1 Scale AI 1 SWE-bench Verified 1 SWE-bench Pro 1 Claude Opus 4.5 1 GLM-5.2 1 GPT-5.5 1
// topics top 5 topics
ai research 1 ai tools 1 large language models 1 developer tools 1 ai safety 1