So happy to defend my PhD thesis and couldn't have done it without a 1 of 1 advisor David Sontag and an incredible committee Eric Horvitz Arvind Satyanarayan Elena Glassman

Congratulations to Dr. Hussein Mozannar. Hussein Mozannar Strong dissertation research exploring multiple dimensions of human-AI collaboration. David Sontag Elena Glassman Arvind Satyanarayan Massachusetts Institute of Technology (MIT)

Co-LLM can help achieve better performance than using a single model alone.

We experimented with co-llm generations at different frequencies for invoking the assistant model, and visualized the generations.

Thanks AK for tweeting our work!

🤝Co-LLM trains a language model to collaboratively decode with other LMs. It decides whether to invoke the other assistant LM for each token. By optimally combining the expertise from each LM, co-llm generation can be better than using…

Learning to Decode Collaboratively with Multiple Language Models

We propose a method to teach multiple large language models (LLM) to collaborate by interleaving their generations at the token level. We model the decision of which LLM generates the next token as a latent

How do LLMs like #ChatGPT impact clinical note readability? LLM augmentations enhance understanding (from user study with 207 participants- 17% increase in understanding for augmentations vs control, p=0.002), but validation was needed to avoid errors: arxiv.org/abs/2401.09637

Help us evaluate how well code LLM models assist humans in completing coding tasks!

Register interest in our online 35 min Python study with $15 compensation forms.gle/BHwJf2b2bp4f4d…

it also serves as paid LeetCode practice!

In my The Boston Globe Op-Ed today, I urge the US government and philanthropists to ensure that Americans can benefit from repurposed generic drugs: bostonglobe.com/2023/12/27/opi….

Reboot Rx #RebootRx #TheSolutionMayAlreadyExist #cancer #drugrepurposing #nonprofit

a hype-free, sober conversation on AI for healthcare by ProfDhar with David Sontag at open.spotify.com/episode/75JIST…

📢I am on the industry and academic job market this year!

My goal is to enable human-AI teams to solve real-world tasks.
I bring a unique skill set that combines ML, HCI, and domain knowledge in healthcare. Please reach out if you have any opportunity that would be a good fit!

We're excited to introduce #INSPECT a large-scale ✨3D multimodal✨medical imaging dataset #NeurIPS2023

19,402 Stanford Medicine Patients
🩻23,248 CT Scans + 📄Paired Radiology Notes
📈Longitudinal EHRs
🩺Clinician-validated task labels
#DataCentricAI #Multimodal #3Dimaging

1/

Our NeurIPS '23 paper by Hussein Mozannar and colleagues at MIT-IBM Watson AI Lab featured on MIT homepage today. Article here --
news.mit.edu/2023/automated…

Monica finished her Ph.D. with me at MIT a year ago and is starting a research group at Duke in Sept 2024. As a leader both in industry Layer Health and in academia, her LLM work is transforming healthcare. Consider applying to Duke CompSci and DukeBiostats

Mike finished his Ph.D. with me at MIT a year ago, and is now recruiting students in CS Johns Hopkins University. Consider applying to work with him - he's a great mentor and top researcher.

AI has uncovered new insights from medical data like disease risk & severity, transforming our understanding of health. But what if AI performance is worse than a human's? Start thinking differently, says #JameelClinic PI David Sontag to ProfDhar on Brave New World podcast 🎧

For medical knowledge tasks, some think that for best results you need fine-tuning and prompts crafted by human medical experts . But no! This study Microsoft Research & Eric Horvitz shows that generalist #GPT4 beats specialized models *and* is less expensive. microsoft.com/en-us/research…

Excited for day 2 of #MLCB2023 today! Tune in at 12:35 EST to catch my talk on #singlecell #foundationmodels ! Great lineup starting at 11:30. Live stream link: youtube.com/watch?v=xhJ-C0…

Verifying LLM outputs is a time-consuming but often necessary endeavor, especially in high-stakes settings.

We propose SymGen, a new way to simplify LLM text verification in settings involving structured data!

📄arxiv.org/abs/2311.09188

Super excited for David Sontag and the Layer Health team! Rare to see such deep ML and deep healthcare talent together.

exclusive: Layer Health inks $4M seed from GV, General Catalyst, Inception Health to unearth data in clinical notes. axios.com/pro/health-tec…