HackerLangs
TopNewTrendsCommentsPastAskShowJobs

mtharrison

no profile record

Submissions

Loupe: Lightweight dev tracing for LLM apps

matt-harrison.com
1 points·by mtharrison·il y a 4 mois·0 comments

Claude Opus 4.6 visible on list models endpoint

2 points·by mtharrison·il y a 5 mois·0 comments

USDA website blames "Radical Left Democrats" for shutdown

usda.gov
30 points·by mtharrison·il y a 9 mois·23 comments

comments

mtharrison
·il y a 12 jours·discuss
Small refinement: the underlying model isn’t stochastic at all. The forward pass is a deterministic function of the weights and input, it just produces a probability distribution over the next token. The stochasticity is an optional sampling step layered on top, not something inherent to LLMs. Greedy/argmax decoding (or temperature 0) makes the whole thing deterministic.

So “purely stochastic” overstates it a bit: the distribution is computed deterministically, and you choose whether to sample from it or not.
mtharrison
·il y a 3 mois·discuss
:)
mtharrison
·il y a 3 mois·discuss
What's in the SpaceX Files and why aren't they already public?
mtharrison
·il y a 8 mois·discuss
Maybe dumb question but how do I just take a sat image and create the scene? The scripts in the repo are all about training which I assume requires you to have the 3d data too.

These sort of projects always look cool but I think the real "wow factor" would be a file upload where you can see the result on your image. I assume there are reasons why this isn't done.
mtharrison
·il y a 9 mois·discuss
The mask becomes the face
mtharrison
·il y a 9 mois·discuss
To what extent is this administration concerned with breaking laws?
mtharrison
·il y a 10 mois·discuss
Strawman Front
mtharrison
·il y a 10 mois·discuss
Is there anyone monitoring US government websites for deletion of pages beyond what people may notice in an ad hoc way? I feel like this should be automated.