To combine the strengths of both models, CSAIL researchers created a sequence model training technique called "Diffusion Forcing." The name comes from "Teacher Forcing," the conventional training scheme that breaks down full sequence generation into the smaller, easier steps of
@mit_csail
-
Diffusion Forcing Enables Robots to Complete Manipulation Tasks Reliably
By
–
By sorting through noisy data & reliably predicting the next steps in a task, Diffusion Forcing can aid a robot in ignoring visual distractions to complete manipulation tasks. It can also generate stable & consistent video sequences and even guide an AI agent through digital
-
MIT’s Diffusion Forcing: Combining Token Prediction and Video Diffusion
By
–
Sequence models have skyrocketed in popularity for their ability to analyze data & predict what to do next.
— MIT CSAIL (@MIT_CSAIL) 17 octobre 2024
MIT’s "Diffusion Forcing" method combines the strengths of next-token prediction (like w/ChatGPT) & video diffusion (like w/Sora), training neural networks to handle… pic.twitter.com/u9CPIGq0tjSequence models have skyrocketed in popularity for their ability to analyze data & predict what to do next. MIT’s "Diffusion Forcing" method combines the strengths of next-token prediction (like w/ChatGPT) & video diffusion (like w/Sora), training neural networks to handle
-
MIT Professor Discusses AI Impact on Chip Design and Moore’s Law
By
–
MIT Prof. on how AI is impacting chip design & the future of Moore’s Law.
— MIT CSAIL (@MIT_CSAIL) 16 octobre 2024
Full video: https://t.co/3gYq4xkjym pic.twitter.com/XlWvX7Q1ZaMIT Prof. on how AI is impacting chip design & the future of Moore’s Law. Full video: https://
bit.ly/3Ss6SHO -
Memory optimization tips for loading large LLMs PyTorch
By
–
Some tips & tricks for reducing how much memory you use when loading larger LLMs & other models in PyTorch: https://
bit.ly/489XajJ v/
@rasbt -
MIT Professor Analyzes LLM Understanding of Complex Financial Concepts
By
–
MIT professor on whether LLMs really understand complex financial concepts.
— MIT CSAIL (@MIT_CSAIL) 9 octobre 2024
Part 2 of the full video: https://t.co/8xofutJ6n6 pic.twitter.com/2tEOI957EJMIT professor on whether LLMs really understand complex financial concepts. Part 2 of the full video: https://
bit.ly/4dAlaO9 -
Ada Lovelace: Celebrating the World’s First Programmer
By
–
In honor of #AdaLovelaceDay, learn about the woman who many view as the world's first programmer: http://
cnet.co/2dZ1Ihp -
Geoffrey Hinton Wins Nobel Physics Prize for AI Research
By
–
Geoffrey Hinton, whose work in neural networks has led many to call him the “Godfather of AI”, co-wins the Nobel Physics Prize for AI research: https://
bit.ly/47X7H1u -
Dell CEO’s 1997 Apple Prediction: A Cautionary Tech Tale
By
–
#otd in 1997, some advice that aged like batteries: "I’d shut [Apple] down and give the money back to shareholders" — Dell CEO Michael Dell: https://
cnet.co/3ZL55lf (v/
@CNET
) #tdih -
Linux Kernel Public Release Anniversary 1991
By
–
#otd in 1991 the Linux kernel was publicly released by Helsinki grad student Linus Torvalds: https://
bit.ly/3N5sfLG