🚨 Excited to attend #ICML2025 and share our latest work (@OptML_MSU) on LLM unlearning -- think of it as AI surgery: removing harmful knowledge while preserving general utility. Catch us at:
🔹 [Paper 1] Tues, July 15 @ 4:30pm PT | E-1108
📄 Invariance Makes LLM Unlearning…
There is obviously way more to Associative Memory than can fit in this thread
On July 14 @DimaKrotov@p_ram_p and I will present a tutorial at #ICML2025 on what we've learned about AMs
Join if you want to be convinced that AI needs physics to reach its full potential 🤙
@Ben_Hoov, @p_ram_p, @baophamhq, and I prepared a Tutorial on Associative Memory for @icmlconf, which we will present next week. (Hopefully) an approachable introduction to the field for the newcomers with hands-on notebooks and some suggested problem sets. In the next few days…
What is Associative Memory and how can it be used in modern AI?
🔷 If I show you an image of a strawberry, can you remember what it smells like or tastes like?
🔷 Can you name a movie by seeing the emojis below?
🔷 Can you name the gentlemen in the picture without seeing…
I am super excited to announce the call for papers for the New Frontiers in Associative Memories workshop at ICLR 2025. New architectures and algorithms, memory-augmented LLMs, energy-based models, Hopfield networks, associative memory and diffusion, and many other exciting…
Excited to share Dense Associative Memory through the Lens of Random Features accepted to #NeurIPS2024🎉
DenseAMs need new weights for each stored pattern–hurting scalability. Kernel methods let us add memories without adding weights!
Distributed memory for DenseAMs, unlocked🔓
Last year's workshop was awesome -- I learned a ton. Definitely in my top-3 most interesting future directions for AI (totally unbiased... 😇).
Highly recommend watching John Hopfield’s talk from the workshop👇 #AI#HopfieldNetworksyoutube.com/watch?v=3OnjX-…
Last year's workshop was awesome -- I learned a ton. Definitely in my top-3 most interesting future directions for AI (totally unbiased... 😇).
Highly recommend watching John Hopfield’s talk from the workshop👇 #AI#HopfieldNetworksyoutube.com/watch?v=3OnjX-…
There has been many discussions around Hopfield Networks of associative memory in the past couple of weeks. But what are the modern frontiers and recent developments in this field? If you want to learn more about this, please check out the NeurIPS 2023 workshop dedicated to this…
Diffusion models are uncannily similar to Associative Memories.
Both “retrieve memories" from an initial signal by descending the energy gradient. Diffusion Models assume an energy. Associative Memories define it.
See our new paper: Memory in Plain Sight arxiv.org/abs/2309.16750
Please consider submitting your best work to the NeurIPS workshop Associative Memory & Hopfield Networks in 2023! Memory-inspired architectures, energy based models, NeuroAI, and much more.
Call for papers: amhn.vizhub.ai/cfp/
Deadline: October 6
amhn.vizhub.ai
Check out this amazing interactive demo of Associative Memory by @Ben_Hoov and @hen_str 👏👏
There is also a mathematical explanation of the underlying mechanics
amhn.vizhub.ai/demo/#demo-ana…
Check out this amazing interactive demo of Associative Memory by @Ben_Hoov and @hen_str 👏👏
There is also a mathematical explanation of the underlying mechanics
amhn.vizhub.ai/demo/#demo-ana…
I am excited to announce Associative Memory & Hopfield Networks Workshop @NeurIPSConf 2023! We have a stellar lineup of invited speakers. The call for contributed papers is now open. See you in New Orleans!
Website: amhn.vizhub.ai
Submission: openreview.net/group?id=NeurI…
K-means clustering requires non-differentiable cluster assignment. Deep clustering replaces discrete assignment with soft assignment (80% of the data point➡️cluster 1, 20%➡️cluster 2). This violates the fundamental premise of clustering – each point belongs to only one cluster.
K-means clustering requires non-differentiable cluster assignment. Deep clustering replaces discrete assignment with soft assignment (80% of the data point➡️cluster 1, 20%➡️cluster 2). This violates the fundamental premise of clustering – each point belongs to only one cluster.
In our #ICML2023 work, @DimaKrotov, @mj_zaki, @p_ram_p, and I uncover a novel connection between Dense Associative Memory dynamics and the inherent discrete assignment to propose an unconstrained continuous relaxation of the discrete clustering problem (ClAM).
Come talk to Pari and me on Tuesday May 2 from 5:00-6:30 pm (Central Africa Time). We'll be happy to talk about anything related to doing research in industry, optimization, generalization, trustworthy AI, and related topics.
Come talk to Pari and me on Tuesday May 2 from 5:00-6:30 pm (Central Africa Time). We'll be happy to talk about anything related to doing research in industry, optimization, generalization, trustworthy AI, and related topics.
6K Followers 2K FollowingCS PhD Student at Stanford Trustworthy AI Research with @sanmikoyejo. Prev interned/worked @ Meta, Google, MIT, Harvard, Uber, UCL, UC Davis
1K Followers 527 FollowingDoing cybernetics (without being allowed to call it that). Assistant Professor @BrownUniversity. Prev: @IBMResearch @MIT, @RutgersU
3K Followers 1K FollowingTheoretical neuroscience, theory of neural computation, physics of learning and intelligence. Assistant Professor of Applied Mathematics @Harvard SEAS
3K Followers 898 FollowingProfessor for CS at the Tuebingen AI Center @uni_tue and affiliated Professor at MIT-IBM Watson AI lab @MITIBMLab - Multimodal learning and video understanding
3K Followers 920 FollowingPrincipal research scientist@IBM Research & Chief Scientist@RPI-IBM AI Research Collaboration & PI@MIT-IBM AI Lab. IJCAI Computers & Thought Award Winner.
1K Followers 3K FollowingTheoretical computational neuroscience postdoc at Cambridge University (CBL lab); PhD from the Hebrew University; also @uricohen42.bsky.social
5K Followers 3K FollowingOfficial account for Emory College of Arts & Sciences. Our faculty, students and alumni are committed to discovery, creativity and academic excellence.
1K Followers 3K FollowingTheoretical computational neuroscience postdoc at Cambridge University (CBL lab); PhD from the Hebrew University; also @uricohen42.bsky.social
1K Followers 527 FollowingDoing cybernetics (without being allowed to call it that). Assistant Professor @BrownUniversity. Prev: @IBMResearch @MIT, @RutgersU
6K Followers 2K FollowingCS PhD Student at Stanford Trustworthy AI Research with @sanmikoyejo. Prev interned/worked @ Meta, Google, MIT, Harvard, Uber, UCL, UC Davis
139K Followers 36 FollowingVancouver, Dec 10-15, 24. https://t.co/ga8aOw6yUO Tweets to this account are not monitored. Please send feedback to [email protected].
3K Followers 1K FollowingTheoretical neuroscience, theory of neural computation, physics of learning and intelligence. Assistant Professor of Applied Mathematics @Harvard SEAS
3K Followers 898 FollowingProfessor for CS at the Tuebingen AI Center @uni_tue and affiliated Professor at MIT-IBM Watson AI lab @MITIBMLab - Multimodal learning and video understanding
3K Followers 640 FollowingI wrote a book.
Free pdf: https://t.co/rFFL7mySnS
Paperback: https://t.co/lF0IgC5T9z
Tweets are my own and don't necessarily represent IBM.
3K Followers 920 FollowingPrincipal research scientist@IBM Research & Chief Scientist@RPI-IBM AI Research Collaboration & PI@MIT-IBM AI Lab. IJCAI Computers & Thought Award Winner.