Excited to share that our latest work has been accepted for NeurIPS publication! We developed a nonlinear state-space model of neural activity and behaviour, Targeted Neural Dynamical Modelling (TNDM): arxiv.org/abs/2110.14853#

Exciting new work on nonlinear state-space modelling of neural activity and behaviour from Cole Hurwitz and folks at @InfAtEd and MIT-IBM Watson AI Lab

Looking forward to seeing everyone at #NeurIPS2021! Our group is hiring both researchers and software engineers. If you are interested in writing code and building software dev tools using deep learning, please reach out!

We are hiring research interns (summer 2022, PhD students only) MIT-IBM Watson AI Lab to work in areas such as contrastive learning, density ratio estimation, generative models and tabular data synthesis. Please DM me your CV if you are interested. #internship

We are hiring a postdoc/research associate at ITUkbh and @mitibmlab (funded by the Climate Change AI grant) to work at the intersection of deep learning and OR towards developing data-driven methods for making the last-mile logistics, greener. Apply here: greenlastmile.ai/jobs

🚨 research position 🚨 work on data-driven tools for greener last-mile logistics 🚲🌳 funded by Climate Change AI, collaboration between ITUkbh, MIT-IBM Watson AI Lab, and Kale AI. in Copenhagen or remotely for up to a year, details here: greenlastmile.ai/jobs

The TuringElect team has been working hard to improve our electoral forecast! If you are interested, you can take a look at some of the results here: withdata.io/election/previ…

Very excited to share my first #ICML2022 paper "Robin Hood and Matthew Effects: Differential Privacy Has Disparate Impact on Synthetic Data." we have a final version: arxiv.org/abs/2109.11429 a summary of our findings: 1/8 🧵

New work from MIT-IBM Watson AI Lab researchers on large scale alignment of LLMs. Check out the models at HF huggingface.co/ibm/merlinite-…

Introducing a community-driven approach to #GenAI model development with the InstructLab project. Help us build #AI in the open: red.ht/4bjRjbR #RHSummit #InstructLab IBM

When people write code, they decompose a big problem into smaller subtasks. Can deep networks do the same? A: Yes! See Kensen Shi talk about our #ICLR2024 paper on Oral 6A Thursday 3:45pm! Paper: arxiv.org/abs/2307.13883 Thread: Below! 🧵

ghostty.org 👻

.Red Hat AI Innovation team just dropped a new research paper on inference-time scaling! 🚨 All built on vLLM. Paper and code here: …abilistic-inference-scaling.github.io Cheers to paper authors Akash Srivastava, Kai Xu, GX Xu, Shivchander Sudalairaj, and Isha Puri!

[1/x] can we scale small, open LMs to o1 level? Using classical probabilistic inference methods, YES! Joint MIT CSAIL / Red Hat AI Innovation Team work introduces a particle filtering approach to scaling inference w/o any training! check out …abilistic-inference-scaling.github.io