Bao Pham
@baophamhq
ML PhD Student at @rpi. Interested in unnormalized probabilistic and modern Hopfield-related models.
Diffusion models create novel images, but they can also memorize samples from the training set. How do they blend stored features to synthesize novel patterns? Our new work shows that diffusion models behave like Dense Associative Memory: in the low training data regime (number…
📢 𝐂𝐚𝐥𝐥 𝐟𝐨𝐫 𝐏𝐚𝐩𝐞𝐫𝐬 – 𝐌𝐞𝐦𝐕𝐢𝐬 @ ICCV | 𝐇𝐨𝐧𝐨𝐥𝐮𝐥𝐮, 𝐇𝐚𝐰𝐚𝐢𝐢 🌺 Topics: 🧠 Memory-augmented models 🎥 Temporal & long-context vision 🤖 Multimodal & scalable systems and more on 𝐦𝐞𝐦𝐨𝐫𝐲 + 𝐯𝐢𝐬𝐢𝐨𝐧 ... 👉OpenReview Submission:…
How to build a factual but creative system? It is a question surrounding memory and creativity in modern ML systems. My colleagues from @IBMResearch and @MITIBMLab are hosting the @MemVis_ICCV25 workshop at #ICCV2025, which explores the intersection between memory and generative…
Consistency Variational Autoencoders (CoVAE) follow naturally from β-VAEs. A family of β-VAEs (with increasing β) can be organized as a sequence of latent encodings with decreasing SNR . This implicit definition of a 'forward process' is used to define a consistency-style loss!
New in the #DeeperLearningBlog: #KempnerInstitute researchers @WangBinxu and @johnjvastola explain their work uncovering the linear Gaussian structure in diffusion models and the potential to use it to enhance performance. bit.ly/4lCauDv #AI #DiffusionModels
1/3) I am biased, but I think this is going to be big! CoVAE: Consistency Training of Variational Autoencoders We unify consistency models with VAEs to obtain a powerful and elegant generative autoencoder! The brainchild of the brilliant @gisilvs (who is looking for jobs!)
Thanks everyone who came to our Tutorial yesterday. It was fun! I will host an additional Q&A session at the IBM Research booth in the West Exhibition Hall A today between 4.30pm and 6pm. If you want to chat about Associative Memory, Energy Transformers, diffusion models, AI &…
Excited to be at ICML to present four papers and recruit new faculty for UMass Amherst! We're hiring in generative AI, NLP, and 3D vision—please feel free to reach out if you're interested!
Adding on to this #ICML2025 tutorial on Associative memories, here is a thread about recent work on simultaneously performing memorization and generalization with Dense Associative Memories
@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…
Energy-based modeling is keen on learning a target data distribution. But, by using the rules of modern Hopfield networks, it can be used to design novel dynamical neural systems whose dynamics are dictated by a global energy function operating in a latent space. For example, we…
During training, diffusion models are being taught to be effective denoisers, like Associative Memory systems. At what point do these models stop being denoisers and behaving like data generators? To learn about how these models arise from being Associative Memory systems to…
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
We are raising $20k (which amounts to $800 per person), to cover their travel and lodging to Kigali, Rwanda in August, from either Nigeria or Ghana. Donate what you can here! donorbox.org/mlc-nigeria-de…
In physics there is an elegant method for computing the correlation functions called generating function. The idea is simple - instead of computing correlators one by one - you define a function of a parameter and compute the average of that new function. Individual correlators…