#Tag · bonfire.cafe

Join us on November 30 from 4:00 pm to 5:00 pm CET (GMT+1) for a talk on using #Generative #AI in #R with Sharon Machlis @smach — Tech Journalist and Data Professional.

In this session, Sharon will offer an accessible, high-level overview of what’s possible today — including how #large #language #models can help you write and improve your R code, as well as add #AI-driven features to your applications.

RSVP 🔗 https://www.meetup.com/rladies-paris/events/312083756/.

#RStats #RStatsFr #Rprogramming #RLadies #Paris

brendan (#39C3に行こう)

@bnys@lasersword.club · 2 weeks ago

looking at gundam models and this rx-78 unleashed just makes me think of karl havoc...too much fucking shit on it #models #gundam #hobby

meme of karl havoc from i think you should leave that says "there's too much fucking shit on me"

image of a gundam model kit that has a billion flaps and hatches open, is overly detailed, its just a lot trust me

R-Ladies Paris

@rladies_paris@mastodon.social · 2 weeks ago

Join us on November 30 from 4:00 pm to 5:00 pm CET (GMT+1) for a talk on using #Generative #AI in #R with Sharon Machlis @smach — Tech Journalist and Data Professional.

RSVP 🔗 https://www.meetup.com/rladies-paris/events/312083756/.

#RStats #RStatsFr #Rprogramming #RLadies #Paris

Hacker News

@h4ckernews@mastodon.social · 3 weeks ago

Multimodal Diffusion Language Models for Thinking-Aware Editing and Generation

https://github.com/tyfeld/MMaDA-Parallel

#HackerNews #Multimodal #Diffusion #Language #Models #ThinkingAware #Editing #Generation #MMaDA

Ulrike Hahn boosted

Fabrizio Musacchio

@pixeltracker@sigmoid.social · 3 weeks ago

🧠 New paper by Deistler et al: #JAXLEY: differentiable #simulation for large-scale training of detailed #biophysical #models of #NeuralDynamics.

They present a #differentiable #GPU accelerated #simulator that trains #morphologically detailed biophysical #neuron models with #GradientDescent. JAXLEY fits intracellular #voltage and #calcium data, scales to 1000s of compartments, trains biophys. #RNNs on #WorkingMemory tasks & even solves #MNIST.

🌍 https://doi.org/10.1038/s41592-025-02895-w

#Neuroscience #CompNeuro

Jaxley: differentiable simulation enables large-scale training of detailed biophysical models of neural dynamics

Fig. 1: Differentiable simulation enables training biophysical neuron models.

Fabrizio Musacchio

@pixeltracker@sigmoid.social · 3 weeks ago

🧠 New paper by Deistler et al: #JAXLEY: differentiable #simulation for large-scale training of detailed #biophysical #models of #NeuralDynamics.

🌍 https://doi.org/10.1038/s41592-025-02895-w

#Neuroscience #CompNeuro

Jaxley: differentiable simulation enables large-scale training of detailed biophysical models of neural dynamics

Hacker News

@h4ckernews@mastodon.social · 3 weeks ago

Continuous Autoregressive Language Models

https://arxiv.org/abs/2510.27688

#HackerNews #Continuous #Autoregressive #Language #Models #NaturalLanguageProcessing #AI #Research #MachineLearning #TransformerModels

arXiv.org

Continuous Autoregressive Language Models

The efficiency of large language models (LLMs) is fundamentally limited by their sequential, token-by-token generation process. We argue that overcoming this bottleneck requires a new design axis for LLM scaling: increasing the semantic bandwidth of each generative step. To this end, we introduce Continuous Autoregressive Language Models (CALM), a paradigm shift from discrete next-token prediction to continuous next-vector prediction. CALM uses a high-fidelity autoencoder to compress a chunk of K tokens into a single continuous vector, from which the original tokens can be reconstructed with over 99.9\% accuracy. This allows us to model language as a sequence of continuous vectors instead of discrete tokens, which reduces the number of generative steps by a factor of K. The paradigm shift necessitates a new modeling toolkit; therefore, we develop a comprehensive likelihood-free framework that enables robust training, evaluation, and controllable sampling in the continuous domain. Experiments show that CALM significantly improves the performance-compute trade-off, achieving the performance of strong discrete baselines at a significantly lower computational cost. More importantly, these findings establish next-vector prediction as a powerful and scalable pathway towards ultra-efficient language models. Code: https://github.com/shaochenze/calm. Project: https://shaochenze.github.io/blog/2025/CALM.

Pin-Ups Of Yore.

@Tattooed_Hippie@mastodon.social · 4 weeks ago

Hacker News

@h4ckernews@mastodon.social · last month

GM Deprecating In-Car App Store for Models as Recent as 2020

https://gmauthority.com/blog/2025/11/these-gm-vehicles-can-no-longer-download-apps-through-their-infotainment-system/

#HackerNews #GM #Deprecating #In-Car #App #Store #for #Models #as #Recent #as #2020

GM #Deprecation #InCarApps #Infotainment #System #2020Models #TechnologyNews

alcinnz boosted

Benedikt Frank

@bene@troet.cafe · last month

The Youtube algorithm recently got me hooked on videos about making SciFi spaceship miniature models, dioramas, kitbashing, movie prop building, practical effects and sculpting weird aliens, monsters and kaiju from plastic toys and garbage. And I love it!
Now I wonder who on mastodon I can follow, to see more this beautiful weird type of art in my feed.
Any recommendations?

#miniatures #models #kitbash #scifi #diorama #props #modelmaking #kaiju #sculpture #diy #maker #bigbotbash #spaceship