sigmoid.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A social space for people researching, working with, or just interested in AI!

Server stats:

596
active users

#probprog

0 posts0 participants0 posts today
Chad Scherrer<p>I got a Probabilistic Programming starter pack going. Hit me up if you're involved with <a href="https://bsky.brid.gy/hashtag/probprog" rel="nofollow noopener" target="_blank">#probprog</a> R&amp;D and want in! <a href="https://go.bsky.app/JfvubEf" rel="nofollow noopener" target="_blank">go.bsky.app/JfvubEf</a><br><br>RE: <a href="https://bsky.brid.gy/convert/ap/at://did:plc:6ls7x4kw3wsz2opik4wobgkm/app.bsky.graph.starterpack/3lbcssvf7yz2x" rel="nofollow noopener" target="_blank">https://bsky.brid.gy/convert/ap/at://did:plc:6ls7x4kw3wsz2opik4wobgkm/app.bsky.graph.starterpack/3lbcssvf7yz2x</a></p>
ML ⇌ Science Colaboratory<p>Tired of waiting forever for MCMC chains to converge? We experimented with using Pathfinder VI to initialize HMC and get early model diagnostics. <a href="https://mlcolab.org/public-events/faster-bayesian-inference-with-pathfinder" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">mlcolab.org/public-events/fast</span><span class="invisible">er-bayesian-inference-with-pathfinder</span></a> <a href="https://fediscience.org/tags/bayesian" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>bayesian</span></a> <a href="https://fediscience.org/tags/probprog" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>probprog</span></a> <a href="https://fediscience.org/tags/probml" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>probml</span></a></p>
Jialin Lu<p>Updated some new thoughts regarding the TerpreT problem and my naive solution</p><p><a href="https://luxxxlucy.github.io/projects/2021_terpret/index.html" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://</span><span class="ellipsis">luxxxlucy.github.io/projects/2</span><span class="invisible">021_terpret/index.html</span></a></p><p>The original TerpreT paper(<a href="https://arxiv.org/abs/1608.04428" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://</span><span class="">arxiv.org/abs/1608.04428</span><span class="invisible"></span></a>)<br />discussed solving program induction by gradient based optimization(after making the program differentiable by relaxation ).</p><p><a href="https://sigmoid.social/tags/probprog" class="mention hashtag" rel="tag">#<span>probprog</span></a> <a href="https://sigmoid.social/tags/programsynthesis" class="mention hashtag" rel="tag">#<span>programsynthesis</span></a> <a href="https://sigmoid.social/tags/neuralnetwork" class="mention hashtag" rel="tag">#<span>neuralnetwork</span></a> <a href="https://sigmoid.social/tags/deeplearning" class="mention hashtag" rel="tag">#<span>deeplearning</span></a> <a href="https://sigmoid.social/tags/NeuroSymbolic" class="mention hashtag" rel="tag">#<span>NeuroSymbolic</span></a></p>
ArviZ<p>Hello Fediverse! This is the official account for the <a href="https://bayes.club/tags/ArviZ" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ArviZ</span></a> project, providing <a href="https://bayes.club/tags/FOSS" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FOSS</span></a> tools for exploratory analysis of <a href="https://bayes.club/tags/bayesian" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>bayesian</span></a> models!</p><p><a href="https://bayes.club/tags/introduction" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>introduction</span></a> <a href="https://bayes.club/tags/ProbProg" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ProbProg</span></a> <a href="https://bayes.club/tags/stats" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>stats</span></a> <a href="https://bayes.club/tags/python" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>python</span></a> <a href="https://bayes.club/tags/JuliaLang" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>JuliaLang</span></a></p>
Chad Scherrer<p><a href="https://bayes.club/tags/introduction" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>introduction</span></a></p><p>I'm Chad. Hello from <a href="https://bayes.club/tags/Seattle" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Seattle</span></a>! 👋 </p><p>Since this is @fosstodon ... My <a href="https://bayes.club/tags/FOSS" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FOSS</span></a> work is in <a href="https://bayes.club/tags/julialang" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>julialang</span></a>, mostly around <a href="https://bayes.club/tags/Bayesian" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Bayesian</span></a> modeling and probabilistic programming (<a href="https://bayes.club/tags/probprog" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>probprog</span></a>)</p><p>This started with Soss.jl, a probabilistic programming language (<a href="https://bayes.club/tags/PPL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PPL</span></a>). I eventually realized I needed primitives with better composability, and started work on MeasureTheory:<br><a href="https://github.com/cscherrer/MeasureTheory.jl" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">github.com/cscherrer/MeasureTh</span><span class="invisible">eory.jl</span></a></p><p>That's coming along well - next it's back to PPL, now with Tilde.jl, similar to Soss but a bit more flexible</p>
ML ⇌ Science ColaboratoryIntroduction
Seth Axen 🪓 :julia:Introduction
Guy Van den Broeck<p>Hi all, my <a href="https://sigmoid.social/tags/introduction" class="mention hashtag" rel="tag">#<span>introduction</span></a>: <br />I&#39;m a prof at <a href="https://sigmoid.social/tags/UCLA" class="mention hashtag" rel="tag">#<span>UCLA</span></a> CS, living in <a href="https://sigmoid.social/tags/LosAngeles" class="mention hashtag" rel="tag">#<span>LosAngeles</span></a>, and researching <a href="https://sigmoid.social/tags/ArtificialIntelligence" class="mention hashtag" rel="tag">#<span>ArtificialIntelligence</span></a>.</p><p>I enjoy bridging <a href="https://sigmoid.social/tags/machinelearning" class="mention hashtag" rel="tag">#<span>machinelearning</span></a> with probabilistic and logical <a href="https://sigmoid.social/tags/reasoning" class="mention hashtag" rel="tag">#<span>reasoning</span></a>.<br />That makes me work on probabilistic programming (<a href="https://sigmoid.social/tags/probprog" class="mention hashtag" rel="tag">#<span>probprog</span></a>), tractable probabilistic models (e.g., <a href="https://sigmoid.social/tags/probcircuit" class="mention hashtag" rel="tag">#<span>probcircuit</span></a>), and <a href="https://sigmoid.social/tags/neurosymbolic" class="mention hashtag" rel="tag">#<span>neurosymbolic</span></a> <a href="https://sigmoid.social/tags/AI" class="mention hashtag" rel="tag">#<span>AI</span></a>.</p><p>Looking forward to some more authentic discourse about AI on this platform.</p>