Reaction
LLMs as monotone compression: a Kolmogorov complexity claim
A post says whether “monotone” is necessary hinges on whether one-way functions exist.
TLDR
A post presents “LLMs = monotone compression” as an answer to why Kolmogorov complexity matters. It says whether “monotone” is necessary hinges on whether one-way functions exist, and links to an arXiv PDF.
Combined views
—
2 Sources, first seen ago
