📢 I am recruiting Ph.D. students for my new lab at New York University! Please apply, if you want to work on understanding deep learning and large models, and do a Ph.D. in the most exciting city on earth. Details on my website: izmailovpavel.github.io. Please spread the word!

🧵🎉 Our new preprint is up, and we’d love your feedback! We're "Getting Aligned on Representational Alignment" - the degree to which internal representations of different (biological & artificial) information processing systems agree. 🧠🤖🔬🔍 #CognitiveScience #Neuroscience #AI

I found this to be very informative for LLM training. the science was just super well done. highly recommended for anyone training transformer based LLMs.

I'm hiring for my performance optimization team at Anthropic! Join our excellent team doing kernels, distributed parallelization, and architecture co-design for GPUs, TPUs and Trainium. No problem if you've only done CPUs before! 🧵 More about us:

If you are curious about increasing representational alignment between neural network and human object similarity, then we have something for you. I won’t be attending #NeurIPS2023 but Andrew Lampinen and Simon Kornblith will be presenting our work in Wednesday‘s morning session 🧠🤖

🔥The CfP for Re-Align is up 🔥 If you are broadly interested in the alignment of two or more representation spaces, consider submitting to our workshop at #ICLR2024 and join us in beautiful Vienna!

Hiring Research Scientists within Google DeepMind - Toronto to join our team & advance the next generation of medical AI, develop cutting-edge LLMs & Multi-modal models to tackle real-world healthcare challenges. Please submit your interest through: forms.gle/2cSbBotUwSfVfu…

⏰Only 10 days left to submit!⏰ The Re-Align Workshop is coming to #ICLR2024 🧠🤖🔍 Come share your representational alignment work at our interdisciplinary workshop at ICLR 2025: representational-alignment.github.io Submission deadline is Feb 2nd!

Have you ever done a dense grid search over neural network hyperparameters? Like a *really dense* grid search? It looks like this (!!). Blueish colors correspond to hyperparameters for which training converges, redish colors to hyperparameters for which training diverges.

I don't have a SoundCloud, but I did join Anthropic last week, and so far it has exceeded my (high) expectations. I would strongly recommend working there (and using Claude). *this project not done at Anthropic -- this was recreational machine learning on my own time.

Introducing Ideogram 1.0: the most advanced text-to-image model, now available on ideogram.ai! This offers state-of-the-art text rendering, unprecedented photorealism, exceptional prompt adherence, and a new feature called Magic Prompt to help with prompting.

Is Ideogram using SD? No. We have Jonathan Ho who came up with denoising diffusion and William Chan (陳俊樂) and Chitwan Saharia who led text to image and text to video at Google. We built everything from scratch, and we have a track record in foundational AI research that powers this

Regular reminder. As an academic, student, postdoc, or professor, you accept less pay and more stress. It is only worth doing if what you do truly matters. Do that high-risk project. Because the boring project does not justify the downsides of academia. For the love of ideas.

I'm excited to join Anthropic to continue the superalignment mission! My new team will work on scalable oversight, weak-to-strong generalization, and automated alignment research. If you're interested in joining, my dms are open.

What aspects of human knowledge are vision models missing, and can we align them with human knowledge to improve their performance and robustness on cognitive and ML tasks? Excited to share this new work led by Lukas Muttenthaler! 1/10

What happens when models see the world as humans do? In our #NeurIPS2024 paper we show that aligning to human perceptual preferences can *improve* general-purpose representations! 📝: arxiv.org/abs/2410.10817 🌐: percep-align.github.io 💻: github.com/ssundaram21/dr… (1/n)

New paper in Neuron Cell Press. Huge thanks to all contributors - Haoran Xu Asa Farahani Simon Kornblith John Duncan and Robert Desimone. Functional architecture of cerebral cortex during naturalistic movie watching doi.org/10.1016/j.neur…

🚨New preprint about the consistency of representational similarity is out! We systematically analyze how representational similarities between vision model representations generalize across datasets and pinpoint their relationship to downstream task behavior. More in the🧵below!