3 ms·
The claim is that the RSI operation is just finding a fixed point of improvement, RSI(LLM) = RSI(LLM) -- for an optimal LLM* which is a fixed point of RSI As
by mjburgess 27d ago
The claim is that the RSI operation is just finding a fixed point of improvement,
RSI(LLM) = RSI(LLM) -- for an optimal LLM* which is a fixed point of RSI
As for eigenvalues/vectors, they're fixed points of (1/val)A or A*val
- ekidd 27d agoEigenvectors represent fixed directions, not fixed magnitudes. From Wikipedia: > More precisely, an eigenvector v of a linear transformation T is scaled by a constant factor lambda when the linear transformation is applied to it: Tv = lambda v . In other words, repeated multiplication of an eigenvector by a matrix can still create exponential growth.
- mjburgess 25d ago> of (1/val)A