Pinned
The new official paper/poster browser for ICLR 2025 is out. Enjoy! iclr2025.vizhub.ai #iclr #ai #iclr2025
Hendrik Strobelt
@henstr.bsky.social
Senior Research Scientist at IBM Research and Explainability lead at the MIT-IBM AI Lab in Cambridge, MA. Interested in all things (X)AI, NLP, Visualization. Hobbies: Social chair at #NeurIPS, MiniConf, Mementor-- http://hendrik.strobelt.com
1.1K followers113 following31 posts
Pinned
The new official paper/poster browser for ICLR 2025 is out. Enjoy! iclr2025.vizhub.ai #iclr #ai #iclr2025
Sunday, Nov 2, 9am !!! Happy to take part at a panel at the coolest workshop* at #IEEEVIS. This year the workshop continues to be "unusual" in a very good sense. It includes an un-conf session to gather this great community. Let's meet there ! visxai.io *bias might be possible :)
Here is the official paper visualization for #iccv2025 : iccv2025.vizhub.ai
The official ICML 2025 paper browser is hot off the press! Enjoy using it and discover new papers and clusters of research. icml2025.vizhub.ai
The official paper browser for CVPR 2025 is out. Enjoy discovering new ideas with the visual interface. cvpr2025.vizhub.ai #ML #AI #CVPR #Visualization
It's on! In its 8th year, the original VISxAI workshop wants your contributions for explaining ML/AI/GenAI principles. Please consider to submit. Deadline: July 30 webpage: visxai.io #ML #AI #GenAI #Visualization
I just updated my MiniClip demo code. It's a very quick implementation of an example of XAI for multimodal models (here a CLIP). Feel free to read the little blog post (code and findings): github.com/HendrikStrob...
Hendrik Strobelt reposted
Replying to @angieboggust.bsky.social
Check out Abstraction Alignment at #CHI2025! 📄Paper: arxiv.org/abs/2407.12543 💻Demo: vis.mit.edu/abstraction-... 🎥Video: www.youtube.com/watch?v=cLi9... 🔗Project: vis.mit.edu/pubs/abstrac... With Hyemin (Helen) Bang, @henstr.bsky.social, and @arvind.bsky.social
Hendrik Strobelt reposted
US's 1994 security assurances in exchange for Ukraine becoming non-nuclear: - Respect Ukraine’s independence and sovereignty; - Refrain from threat of force; - Refrain from economic coercion; - Seek immediate UNSC action in case of nuclear threats. treaties.un.org/doc/Publicat...
Hendrik Strobelt reposted
Question about @hf.co Huggingface Spaces. I do have an existing code base that I would like to push to a space, but there is no way of creating an empty repo space. HELP !!!
I have worked with a lot of frameworks for front-ends, but I have never experienced something something as BAD as @gradio-hf.bsky.social -- opinionated for only simple use cases, dead component ecosystem, and creating you own components takes major investment. I am forced to use it for HF spaces. 😡
Hendrik Strobelt reposted
Hendrik Strobelt reposted
Hendrik Strobelt reposted
DeepSeek R1 shows how important it is to be studying the internals of reasoning models. Try our code: Here @canrager.bsky.social shows a method for auditing AI bias by probing the internal monologue. dsthoughts.baulab.info I'd be interested in your thoughts.
Hendrik Strobelt reposted
It was inevitable, but it’s still good to be aware that this is happening here, now. If you see suspicious account/comment, remember: Don’t engage. Don’t argue. Don’t clap back. Don’t screenshot and share. Just block and move on. today.rtl.lu/news/busines...
>13k views… woohoo, I hope it was helpful to those who did use it.
🎺 Here comes the official 2024 NeurIPS paper browser: - browse all NeurIPS papers in a visual way - select clusters of interest and get cluster summary - ZOOOOM in - filter by human assigned keywords - filter by substring (authors, titles) neurips2024.vizhub.ai #neurips by IBM Research Cambridge
Hendrik Strobelt reposted
Hendrik Strobelt reposted
Apply to the NDIF Summer Engineering Fellowship to come to Boston to work on the National Deep Inference Fabric, a cutting-edge infrastructure for research in large-scale AI. A really interesting project that crosses ML, HPC, Systems, PL, UX/Viz, research, and engineering. ndif.us/fellowship.h...
More big news! Applications are open for the NDIF Summer Engineering Fellowship—an opportunity to work on cutting-edge AI research infrastructure this summer in Boston! 🚀
Most innovative paper at #NeurIPS imho. Can we create a network that becomes the physical chip instead of running on a chip? Inference speedups and energy preservation are through the roof ! Oral on Friday at 10am PT neurips.cc/virtual/2024...
Hendrik Strobelt reposted
Excited to share our #NeurIPS 2024 Oral, Convolutional Differentiable Logic Gate Networks, leading to a range of inference efficiency records, including inference in only 4 nanoseconds 🏎️. We reduce model sizes by factors of 29x-61x over the SOTA. Paper: arxiv.org/abs/2411.04732
🎺 Here comes the official 2024 NeurIPS paper browser: - browse all NeurIPS papers in a visual way - select clusters of interest and get cluster summary - ZOOOOM in - filter by human assigned keywords - filter by substring (authors, titles) neurips2024.vizhub.ai #neurips by IBM Research Cambridge
with @bhoov.bsky.social and data from Lee Campbell new features to come soon: - export a selection of papers as CSV
[Recent History] LSTMVis (2017) has been one of my first projects in the VIS+ML field. It tries to help identify meaningful state change patterns in RNN hidden states and map them to NLP concepts. #NLP #XAI #VIS lstmvis.vizhub.ai vimeo.com/172145096
It was also my very first collaboration with the dream team around @sebgehr.bsky.social , @srushnlp.bsky.social , and Hanspeter Pfister.
Hendrik Strobelt reposted
Replying to @lchoshen.bsky.social
Of course, babies only take small naps See you at BabyLM Workshop in EMNLP25 Suzhou More details soon here, on slack and on babylm.github.io Until then, please share your suggestions for next year
Oops — just seeing quite some friendly faces here. Thanks @domoritz.de for bringing attention to blueksy.
You are up to date.