clern 😳 (@_buggingface_) header
clern 😳 (@_buggingface_) profile picture

goofy goober

Brooklyn, NY
Joined May 2023
Photo by clern 😳 (@_buggingface_)Photo by clern 😳 (@_buggingface_)Photo by clern 😳 (@_buggingface_)Photo by clern 😳 (@_buggingface_)Photo by clern 😳 (@_buggingface_)Photo by clern 😳 (@_buggingface_)
clern 😳 retweeted
Datology Curation Studio is out today, along with the experiments I spent all summer running to test it. I'm excited to finally talk about the research engineering that went into them! A lot of it was new for us, and the whole team really came together to make it work. 🧵
Today, we’re excited to launch the Datology Curation Studio, customizable frontier data curation for every team training its own models. Data quality is the ultimate compute multiplier, and Curation Studio makes it easy for everyone.
1
10
3
36
1,454
clern 😳 retweeted
Jeffrey has been on a mission, investigating and developing local learning rules. We know that the brain can't do backprop - at least not exactly - so it would be amazing if we could figure out what it is doing instead.
Introducing PC-ALM, a local-learning alternative to backpropagation. Our method trains 1000-layer neural nets using only local dynamics, and without backprop. Blog: pub.sakana.ai/pc-alm/ Standard deep learning relies on backpropagation. The brain, however, cannot implement backpropagation, at least not exactly. How can a physical system, such as the brain, solve multilayer credit assignment without explicit use of backprop? We look for inspiration in two related fields: distributed optimization and NeuroAI. In NeuroAI, predictive coding asks each neuron activation to solve an energy-based inference problem instead of using a standard forward pass. That inference step can be implemented as energy-minimization dynamics on local prediction errors. This perspective -- each layer as a dynamical system -- has proven promising, but performance of predictive coding hasn't scaled well with depth. Credit signals at far ends of the network struggle to diffuse into internal layers. We turn to distributed optimization, generalizing predictive coding to use an augmented Lagrangian instead of energy. This motivation stems back to a classic 1988 paper by LeCun, showing that the Lagrange multipliers of a deep network can be identified with gradients of a supervised loss. The augmented Lagrangian then bridges LeCun's perspective to the standard predictive coding that is used in NeuroAI. We find that this new perspective yields a natural PC-like alternative to backpropagation, resulting in a method we call PC-ALM. PC-ALM differs from PC in that it introduces dual neurons (Lagrange multipliers) as part of the layer-local dynamics, resulting in each layer acting as a PI feedback control system to minimize local prediction errors. We find that PC-ALM is capable of propagating signals to seemingly arbitrary depth, especially in deep narrow networks where standard PC struggles to learn. Ultimately, our motivation here is to understand how distributed physical systems, such as the brain, can compute credit signals using only local coupling and local dynamics. PC-ALM may also inform deep learning in neuromorphic hardware, where dynamics are cheaper than on GPUs. Paper: arxiv.org/abs/2605.31022 Code: github.com/SakanaAI/pc-alm
5
64
7,331
sharing a new paper! Augmented Lagrangian Predictive Coding I think one of the coolest unsolved problems is how the brain does credit assignment. thread 🧵
Introducing PC-ALM, a local-learning alternative to backpropagation. Our method trains 1000-layer neural nets using only local dynamics, and without backprop. Blog: pub.sakana.ai/pc-alm/ Standard deep learning relies on backpropagation. The brain, however, cannot implement backpropagation, at least not exactly. How can a physical system, such as the brain, solve multilayer credit assignment without explicit use of backprop? We look for inspiration in two related fields: distributed optimization and NeuroAI. In NeuroAI, predictive coding asks each neuron activation to solve an energy-based inference problem instead of using a standard forward pass. That inference step can be implemented as energy-minimization dynamics on local prediction errors. This perspective -- each layer as a dynamical system -- has proven promising, but performance of predictive coding hasn't scaled well with depth. Credit signals at far ends of the network struggle to diffuse into internal layers. We turn to distributed optimization, generalizing predictive coding to use an augmented Lagrangian instead of energy. This motivation stems back to a classic 1988 paper by LeCun, showing that the Lagrange multipliers of a deep network can be identified with gradients of a supervised loss. The augmented Lagrangian then bridges LeCun's perspective to the standard predictive coding that is used in NeuroAI. We find that this new perspective yields a natural PC-like alternative to backpropagation, resulting in a method we call PC-ALM. PC-ALM differs from PC in that it introduces dual neurons (Lagrange multipliers) as part of the layer-local dynamics, resulting in each layer acting as a PI feedback control system to minimize local prediction errors. We find that PC-ALM is capable of propagating signals to seemingly arbitrary depth, especially in deep narrow networks where standard PC struggles to learn. Ultimately, our motivation here is to understand how distributed physical systems, such as the brain, can compute credit signals using only local coupling and local dynamics. PC-ALM may also inform deep learning in neuromorphic hardware, where dynamics are cheaper than on GPUs. Paper: arxiv.org/abs/2605.31022 Code: github.com/SakanaAI/pc-alm
11
40
9
315
34,420
This post blew up in finite time. I’m Navier stoked!
I lowkey think Google might have solved Navier Stokes
1
3
158
clern 😳 retweeted
A really excellent book. A few people independently told me this was one of their favorite books over a decade ago. I bought it, and it became one of the textbooks on my shelf I revisit from time to time to spark the joy of holding ideas to a different light. It brings to life the elegance of information theory. A good day to recognize 10 years from the passing of David McKay.
22
144
11
1,424
66,947