Donal Fellows
dkf@cyberplace.social
<p>Research Software Architect for Research Software Engineering at the University of Manchester.</p><p>Tcl programmer, but also many other techs too. Lots of software automation engineering these days.</p><p>I want real AGI, not fake stuff powered by overtrained LLMs.</p>
Posts
-
Post #2446894
Normally, I don&#39;t share LinkedIn junk... but this time I&#39;ll make an exception: https://www.linkedin.com/posts/congratulations-to-our-research-software-share-7459961130879193088-YKAp?utm_source=share&amp;utm_medium=member_desktop&amp;rcm=ACoAABiy5lEBsYvrxI1O5NkWVzsQjhsnkVw1EvE
-
Post #2029543
Two new features added to the forthcoming Tcl 9.1 today. #1 is a way to pick items out of a list: set bools [lfilter x $theList {string is boolean -strict $x}]
-
Post #1477109
They perform projections in a (very!) high-order smooth differentiable space. For problems which that describes, they work excellently. But when the problem doesn&#39;t fit that pattern, they&#39;re not so great. (The underlying math of hypersurfaces doesn&#39;t require differentiability, but the training algorithm guarantees it.) Public LLMs have an awfully large training dataset. Much bigger than you think.
-
Post #1477108
@rayckeith Don&#39;t think of him as Batman. Think of him as more like Lex Luthor. He even looks the part.