
Cyberneticist @GoogleDeepMind
San Francisco
Joined September 2014
- Tweets82
- Following852
- Followers1.1K
- Likes398
James Harrison retweeted
Deepseek R1 inference in pure JAX! Currently on TPU, with GPU and distilled models in-progress. Features MLA-style attention, expert/tensor parallelism & int8 quantization. Contributions welcome!
James Harrison retweeted
Are you still using hand-designed optimizers?
Tomorrow morning, I’ll explain how we can meta-train learned optimizers that generalize to large unseen tasks!
Don't miss my talk at OPT-2024, Sun 15 Dec 11:15-11:30 a.m. PST, West Ballroom A!
teapawt.net/benjamintherien/status…
Learned optimizers can’t generalize to large unseen tasks…. Until now! Excited to present μLO: Compute-Efficient Meta-Generalization of Learned Optimizers! Don’t miss my talk about it next Sunday at the OPT2024 Neurips Workshop :) 🧵arxiv.org/abs/2406.00153 1/N
New: a neural net-based approach to Bayesian optimization that performs well in both classic, small-scale problems, and can efficiently scale far beyond GP surrogate models.
If you're at NeurIPS, come by our poster at the Bayesian decision-making workshop today!
More info👇
Our results show our approach of VBLLs + MLP features + our training approach yields SOTA or near-SOTA performance on a bunch of problems from classic 2D Ackley, all the way to much more challenging problems like Lunar Lander controller tuning
1
2
211
Our approach can be applied to a very very wide range of problems---we include multi-objective BayesOpt as an example, but we're excited about massively scaling up this approach on a ton of problems.
Paper link here: arxiv.org/pdf/2412.09477
1
169
James Harrison retweeted
Going to @itssieee ITSC'24? Check our tutorial on Data-driven Methods for Network-level Coordination of AMoD Systems
Organized with @DanieleGammelli, Luigi Tresca, Carolin Schmidt, @jmes_harrison, Filipe Rodrigues, Maximilian Schiffer, @drmapavone
rl4amod-itsc24.github.io
Want a really simple (and cheap!) way to improve neural net calibration and get practical epistemic uncertainty estimates?
At ICLR this year: Variational Bayesian Last Layers
Try it out:
1. pip install vbll
2. a couple one line changes to your current training pipeline
🧵
We tested VBLLs in regression, classification, and active decision-making. I'm particularly excited about the bandit performance---VBLLs enable nearly free active learning!
Plus, you can use VBLLs jointly with other Bayesian methods, or add them to your model post-training.
1
4
807
For more info, you can check out:
Paper (spotlight @ ICLR): arxiv.org/abs/2404.11599
Torch implementation + docs + tutorial colabs: github.com/VectorInstitute/v…
JAX implementation: coming soon!
A question we have been thinking about for a long time: what is the natural architecture for a learned optimizer? We now have an important part of the answer---we can automatically construct expressive optimizers based on optimizee network symmetries. Check out Allan's thread!
James Harrison retweeted
📝Quiz time: when you have an unrolled computation graph (see figure below), how would you compute the unrolling parameters' gradients?
If your answer only contains Backprop, now it’s time to add a new method to your gradient estimation toolbox!
Want to learn about learned optimization? I gave a tutorial at @CoLLAs_Conf which is now public! youtu.be/FMjYwthtoN4?si=kiPD…
James Harrison retweeted
Looking forward to getting started at #ICML! Happy to chat about RL, learning-based control, and Graph ML.
Make sure to drop by our poster! (Wed 26 Jul 2 p.m. PDT)
Excited to share that our paper on Graph-Reinforcement Learning was accepted at #ICML2023!
We present a broadly applicable approach to solve graph-structured MDPs through the combination of RL and classical optimization.
Website: sites.google.com/stanford.ed…
🧵👇(1/n)+quoted tweet
Graph deep learning and bi-level RL seem to work exceptionally well for a whole bunch of critically important real-world problems like supply chain control. Plus, it easily combines with standard linear programming planners in OR. Check out @DanieleGammelli's thread for info!
Excited to share that our paper on Graph-Reinforcement Learning was accepted at #ICML2023!
We present a broadly applicable approach to solve graph-structured MDPs through the combination of RL and classical optimization.
Website: sites.google.com/stanford.ed…
🧵👇(1/n)+quoted tweet
James Harrison retweeted
Happy to share that our latest work on adaptive behavior prediction models with @jmes_harrison @GoogleAI and @drmapavone @NVIDIAAI has been accepted to #ICRA2023!
📜: arxiv.org/abs/2209.11820
We've also recently released the code and trained models at github.com/NVlabs/adaptive-p…!!
Really nice + concise VeLO explainer!
Why tune optimizers hyperparameters (ex. Adam) by hand, if one can train a neural network to behave like an optimizer and dynamically find the best update for your neural network’s weights?
In this video, we explain the VeLO learned optimizer!👇
📺 youtu.be/9a6PQJxzUpM




















