ConceptAI & ML
Bfloat16
A 16-bit floating-point format that maintains the same exponent bits and range as FP32 while truncating the mantissa for higher performance without needing gradient scalers.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Exponent and mantissa1Floating-point number representation1
Opens up
Concepts that name this one as a prerequisite.
Lecturers
1Lectures that assume it
1Connections
3The lectures that define it or assume it, and the concepts next to it.