Tag: AI evaluation
Browse our exclusive articles!
SWE-bench Verified: Real-World AI Software Benchmark
Discover SWE-bench Verified, a human-validated AI benchmark for real-world software engineering tasks to improve model reliability and performance.
OpenAI o1 System Card External Testers Acknowledged
Discover how OpenAI's o1 system card external testers contributed to safer, more effective AI through collaboration and transparency.
MLE-bench: Benchmarking AI Agents in ML Engineering
Discover MLE-bench, a benchmark tool to evaluate AI agents on machine learning engineering tasks for improved accuracy, efficiency, and scalability.
SimpleQA: Benchmarking Factual Accuracy in Language Models
Discover SimpleQA, a new benchmark designed to evaluate and improve factual accuracy in AI language models for reliable information.
PaperBench: Benchmarking AI Research Replication Skills
Discover PaperBench, the new benchmark assessing AI's ability to replicate research, ensuring reliability and advancing AI innovation.