iiinova — where your digital self keeps evolving

Reposted by Mononito Goswami

Nari Johnson @narijohnson.bsky.social · 24/06/2026

All AI evaluations embed assumptions about what "good" behavior looks like. Our #FAccT2026 paper explores how we can center the perspectives of impacted communities - specifically, subjects of AI-generated media - in designing LLM-as-a-judge evaluation rubrics.

A flyer teasing our research paper. The flyer contains a visualization of our approach: First, community members participate in creating an evaluation rubric, visualized as a list of criteria. The rubric is then given as input to an LLM judge.

183 0

Mononito Goswami @mononitogoswami.bsky.social · 25/03/2025

📣 Website & Cfp: icml-structured-fm-workshop.github.io Submission deadline: May 19 Acceptance notification: June 9 Workshop date: July 18/19 📄 Page limit: Short papers with up to 4 pages 👀 Misc.: Double blind, Non-archival, Posters & Orals for accepted papers

000 0

Mononito Goswami @mononitogoswami.bsky.social · 25/03/2025

🚀 Interested in time series 📈📉 and tabular data? Then our @icmlconf.bsky.social’25 Workshop on Foundation Models for Structured Data is exactly what you are looking for 😁! Send us your best work and join us in Vancouver 🇨🇦🍁

120 0

Mononito Goswami @mononitogoswami.bsky.social · 13/02/2025

📜 Check out our papers hot off the press! [O1] Investigating Compositional Reasoning in Time Series Foundation Models (arxiv.org/abs/2502.06037) [O2] Exploring Representations and Interventions in Time Series Foundation Models (arxiv.org/abs/2409.12915)

arxiv.org

Investigating Compositional Reasoning in Time Series Foundation Models

Large pre-trained time series foundation models (TSFMs) have demonstrated promising zero-shot performance across a wide range of domains. However, a question remains: Do TSFMs succeed solely by memori...

100 0

Mononito Goswami @mononitogoswami.bsky.social · 13/02/2025

Also we're continuing to push the boundaries of science on time series foundation models, and I am thrilled about our recent work on 🔹 compositional reasoning [O1] and 🔹 steering [O2] time series foundation models!

100 0

Mononito Goswami @mononitogoswami.bsky.social · 13/02/2025

🚀 Excited to share that MOMENT has been downloaded 1.3M+ times on HuggingFace, with over 280K downloads in just the last month! MOMENT is the BERT for time series. It can solve many tasks across different domains, including healthcare! Best part? It's completely open source!

120 0

Mononito Goswami @mononitogoswami.bsky.social · 15/12/2024

Also my awesome collaborators and mentees, Arjun Choudhry and @inverse-hessian.bsky.social are looking for PhD positions. If you are looking for students with experience in multimodal foundation models, efficiency, mechanistic interpretability, I don’t know of anyone better than them!

000 0

Mononito Goswami @mononitogoswami.bsky.social · 15/12/2024

Come chat with me and my wonderful collaborators about Time Series Foundation Models at the NeurIPS Workshops on Time Series Age of Large Models, Systems 2 Reasoning, and Foundation Models Interventions!

130 0

Mononito Goswami @mononitogoswami.bsky.social · 14/12/2024

Interested in learning about making univariate time series foundation models multivariate with only 24 parameters? Check out our poster on Long Context TSFMs in the NeurIPS’24 FITML Workshop, East Exhibition Hall A at 11:30 AM

000 0

Reposted by Mononito Goswami

Mononito Goswami @mononitogoswami.bsky.social · 09/12/2024

Excited to be at NeurIPS'24, where I'll be presenting at several workshops! Looking forward to chatting about (time series & tabular) foundation models, data science agents, or ML for healthcare! Also, I'm also on the industry job market, looking forward to connect 😁!

0225 0

Mononito Goswami @mononitogoswami.bsky.social · 09/12/2024

0225 0

Reposted by Mononito Goswami

Kush Varshney कुश वार्ष्णेय @krvarshney.bsky.social · 07/12/2024

If you’re headed to NeurIPS 2024, and want to learn about IBM Research Human-Centered Trustworthy AI, there are many many opportunities to do so. 1. Start with the official NeurIPS explorer by @henstr.bsky.social and @benhoover.bsky.social. It is infoviz par excellence. neurips2024.vizhub.ai

neurips2024.vizhub.ai

Tips

193 0

Reposted by Mononito Goswami

Vaishnavh Nagarajan @vaishnavh.bsky.social · 03/12/2024

if you're a PhD student at CMU doing AI/ML, lmk if you want to be added to this starter pack. (I don't belong in this list, but I don't know how to remove myself from this pack 😂) go.bsky.app/9APVxQQ

3143 0

Reposted by Mononito Goswami

Ahmad Beirami @abeirami.bsky.social · 27/11/2024

The question that a reviewer should ask themselves is: Does this paper take a gradient step in a promising direction? Is the community better off with this paper published? If the answer is yes, then the recommendation should be to accept.

5649 0

Reposted by Mononito Goswami

Zachary Lipton @zacharylipton.bsky.social · 26/11/2024

Medically adapted foundation models (think Med-*) turn out to be more hot air than hot stuff. Correcting for fatal flaws in evaluation, the current crop are no better on balance than generic foundation models, even on the very tasks for which benefits are claimed. arxiv.org/abs/2411.04118

arxiv.org

Medical Adaptation of Large Language and Vision-Language Models: Are We Making Progress?

Several recent works seek to develop foundation models specifically for medical applications, adapting general-purpose large language models (LLMs) and vision-language models (VLMs) via continued pret...

825857 0