olens_ablations

Model Card

Every arm here differs from the recipe of record by exactly one flag. Data, pool, crop seed, layer set, learning rate, batch and seeds are held fixed across arms so the only free variable is the one named. The whitened-FVE ruler that the program optimises and reports anti-correlates with lens quality. The AR trained without whitening scores far worse on the ruler and makes a materially better lens, and running both to saturation does not rescue it. Compare arms only at MATCHED cumulative steps. Arms stopped early by a budget cut have lower totals, and reading their endpoints against a fully-trained arm inverts the ordering: unitcos looks worse than base on endpoints but is identical to it…

Excerpt from the card by Agam Bhatia.

Identity and Version

Repository

agu18dec/olens_ablations

Publisher

Agam Bhatia

Task

Not stated by the source

Modality

Other

Library

Not stated by the source

Parameters

Not stated by the source

Languages

Not stated by the source

Revision

e9585c4502bf0c633a95e67bdf096a331e43c5ea

First published

2026-09-20

Last updated

2026-09-30

Files and Weights

3,325 files, 424.8 GB in total. The weights are 915 files totalling 424.8 GB in pt, safetensors.