TLDR AI · DEEP DIVES & ANALYSIS
PHI 4 REASONING 100% ON PRIVATE GRADUATE LINEAR ALGEBRA EXAM
Summary from the TLDR AI newsletter
The new synthetically trained reasoning model from Microsoft shows strong local math and code performance, even in the face of poor internal world knowledge.