This is an empirical study of sparse autoencoders addressing a defined open problem, but it reports unexpected null results (no recoveries, no merges) and characterizes a previously undescribed diffuse phase rather than validating the proposed phase diagram; the work is unreviewed and requires confirmation.
PreliminaryAlexis D. Plascencia
Reported
Independent fits evaluated (sampl…200
Grid cells evaluated (subset)10 of 165
Total additional fits (full grid)3,300