Pairwise ELO Rankings Streamline Awards and Performance Reviews
So many judging tasks could be improved by aggregating partial orderings, and in the limit, just ordering pairs. The annual Libertarian Futurist Society novel awards discussion is starting, and while I would like to participate on some level, there is no way I have time to read an entire slate of novels. However, I will likely read at least two from the list, and I could give a relative assessment. This cries out for the use of something like ELO ranking, as in chess competition, perhaps with some suggestions to get sufficient coverage. Peer and out-of-chain employee performance calibrations could probably also benefit from a greater quantity of sparse pairwise comparisons
GPU Power Draw Beats Scheduling as Utilization Metric
Without getting all the way down to performance counters, GPU power from nvidia-smi is a better indicator of true utilization than job scheduling or “gpu busy”. I would love to see animated “heat maps” of the big data centers, with...
SigReg Loss Boosts Stable Pixel‑based World Models
Paper review: LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels https://t.co/TpFFnwPWkc Nice clean github: https://t.co/HOuqEf0HaF This is the application of the LeJEPA results to world models, trained offline on experience from three different robotics style tests with one to two million steps in...
Use Advisors as a Skill‑Based Pool, Not Vibe Checks
The corporate advisory boards that I have been a part of have almost exclusively been “vibe checks”, where presentations are made about work the company is doing, and the advisory panel chats about things for a while. This seems like a...
Epic's $0‑to‑$1M Tax Boosts Dev Activity, Avoids Waste
Companies like Meta subsidize third party developers in various ways to help grow their platforms, then take 30% of the developer revenue right back with the platform tax, which is a wasteful churn. Epic’s tax structure of zero for the first...
Ethics Chapter Misses Depth, Confuses Relevance with Legitimacy
I still give the book Understanding Deep Learning by Simon J.D. Prince a good recommendation, but chapter 21: Deep learning and Ethics was sloppy. It could have been a chapter to really dig in on case studies, but it was...
Low IQ, High Agency: AI Trust Wins
The modern age has richly rewarded people with a combination of high intelligence and high agency. Now that many aspects of intelligence are successfully being automated, it seems likely that people with relatively lower intelligence but exceptional agency will come...
Baseline Agent Outperforms Standard Algorithms in Physical Atari Replication
Another RL team replicated our Physical Atari work and compared my baseline agent against several standard algorithms. https://t.co/uyWjFXxnZO
PyTorch's 64‑bit Index Standard Simplifies Development
Pytorch made the right call standardizing on signed 64 bit indexes. I would probably still be rather pointlessly making case by case decisions to use int32 if it were an option. Some old habits linger.
Will Journals Reject Papers Using Real Code Variable Names?
Would papers get rejected if they used legal code with variable names instead of Greek laden pseudo code?
Online Learning Bridges Performance Gaps Between Identical Robots
Some interesting experiments with our “physical Atari” RL framework — one of our main points was that the flawless simulators didn’t transfer well at all to the messy reality of cameras and servos, but perhaps less obviously, even transferring from...
Weight Decay Softly Prunes Noisy Features, Not Just Simplicity
Weight decay is usually presented as “encouraging simpler solutions”, but I tend to think that the real benefit is the soft pruning of noisy / unhelpful features. Without decay, a weight can random-walk to a large value even if the...
Discover PyTorch’s Pixel_unshuffle: Skip Custom Tensor Hacks
Always a slightly mixed feeling to write pretty good first-principles code to do some tensor rearrangement, only to find that PyTorch has a built in function that does it faster. I had made a point of at least skimming the docs...
Base‑2 Softmax Boosts Fixed‑Point Hardware Efficiency
I wonder if doing softmax in base 2 instead of base e could be a useful optimization on some fixed point hardware. Prepend a 1 to the fractional part and shift left by the integral part as a close-ish approximation...
Consistent Naming Crucial Even With Autocomplete Assistance
It doesn’t matter as much in the age of autocomplete as it used to, but the tiny mental friction of naming inconsistencies like “start_dim” vs “keepdim” still sticks out for me. Carefully pick names as if millions of developers may...
