SoTALab Blog · 2 articles

Research & Notes

Benchmarks, evaluation methodology, and operating notes from the SoTALab team.

DataJun 23, 20262,932 words14 min read

Beyond the Final Answer: Process-Oriented Math Training Data

Answers are easy to verify; reasoning is hard to constrain—five advanced math examples show why process-oriented training data matters more than a question bank.

ResearchJun 8, 20261,429 words7 min read

What Do Leaderboard Scores Actually Measure?

Tasks, data, scoring, and runtime—what benchmark numbers can tell you, what they cannot, and how to read a leaderboard.