Top all-time · #2365
Eric Jang
Personal blog · blog.evjang.comaiprogrammingscience
Eric Jang's personal blog at blog.evjang.com covers ai, programming and science. Since 2016, 31 different Hacker News readers have linked it 54 times, most often “Dijkstra's in Disguise”. The last post was in 2021, but the archive is still widely read.
Technology, A.I., Careers
Most-linked posts?
- Dijkstra's in Disguise6 people · 6 mentions · first linked 2018-08-08
- Eric Jang: Differentiabe Path Tracing on the GPU/TPU4 people · 4 mentions · first linked 2019-11-29
- How to get an internship3 people · 3 mentions · first linked 2016-07-24
- Meta-Learning in 50 Lines of JAX2 people · 2 mentions · first linked 2019-02-21
- Sovereign Arcade: Currency as High-Margin Infrastructure2 people · 2 mentions · first linked 2021-05-27
- Stonks Are What You Can Get Away With: NFTs and Financial Nihilism2 people · 2 mentions · first linked 2021-06-20
- My Internship Experiences at Pixar, Google, and Two Sigma2 people · 2 mentions · first linked 2016-10-24
- Three questions that keep me up at night2 people · 2 mentions · first linked 2020-04-02
- Expressivity, Trainability, and Generalization in Machine Learning2 people · 2 mentions · first linked 2017-11-20
- Aesthetically Pleasing Learning Rates2 people · 2 mentions · first linked 2018-04-01
What Hacker News says?
In the process of learning about a family of algorithms in machine learning I also gained some physical intuition of determinants (same diagram as 3Blue1Brown, but applied in a different context of "squashing and stretching" probability mass): https://blog.evjang.com/2018/01/nf1.html
Comment linking /2018/01/nf1.html
on “How to Understand Things”Submitted Dijkstra's in Disguise as a story · 450 points
HN discussionIf you like graph theory Dijkstra’s in Disguise is a great read: https://blog.evjang.com/2018/08/dijkstras.html Related HN thread: https://news.ycombinator.com/item?id=17745779
Comment linking /2018/08/dijkstras.html
on “Start Doing Graph Theory”Submitted Fun with Snapchat's Gender Swapping Filter as a story · 290 points
HN discussionI appreciate the authors calling out #2. ImageNet and CIFAR are both "solved" benchmarks, to the extent that state of the art algorithms by now are likely overfitting to the specific dataset details. In particular, objects in ImageNet are nearly always in canonical pose, with no occlusion, few confounding objects, illuminated, and in a semantically-obvious configuration. For some applications this is acceptable but…
Comment linking /2019/02/bagnet.html
on “Myths in Machine Learning Research”Submitted Stonks Are What You Can Get Away With: NFTs and Financial Nihilism as a story · 172 points
HN discussionI agree that the sort of sampling bias/distributional shift you mention is definitely a real problem in deep learning. It so happens that yesterday I read a blog post that elaborates on this problem clearer than I can. Although it's still true that allowing arbitrary distributional shift is impossible, it does make sense to allow for distributional shift across the natural manifold (manifold of all "allowable" data)…
Comment linking /2017/11/exp-train-gen.html?spref=tw
on “Mathematics of Deep Learning [pdf]”Submitted Robots Must Be Ephemeralized as a story · 104 points
HN discussionRight.I was focusing more on the CUDA-kernels-on-NPU line of thought from patrikthebold's message and its alternatives, which as you say, is not a reality now either. In the Jax style implementation scenario the compiler part of JAX is better inspiration, maybe along the lines of this case study of a path tracer running on a TPU: https://blog.evjang.com/2019/11/jaxpt.html - I don't think Chapel or Futhark would adop…
Comment linking /2019/11/jaxpt.html
on “Demystifying NPUs: Questions and Answers”Submitted Summary of the Neural Information Processing Systems Conference 2016 as a story · 99 points
HN discussionSelf-Plug: If you're interested in a Jupyter notebook tutorial on implementing a basic differentiable path tracer (a physically based rendering algorithm), I wrote up a tutorial here: https://blog.evjang.com/2019/11/jaxpt.html
Comment linking /2019/11/jaxpt.html
on “Neural Renderer and Differentiable Rendering [video]”Submitted Don't Mess with Backprop: Doubts about Biologically Plausible Deep Learning as a story · 91 points
HN discussionThe superformula depends of four parameters and is able to model many different curves. I wonder if that superformula would be useful to learn to generalize the form of a curve given few points. It could be that, in same way, the four parameters of that curve are a orthogonal bases in the hypothesis space, in the sense that each parameters add a lot the information. If this intuition has any meaning, it could be the…
Comment linking /2017/11/exp-train-gen.html
on “Superformula”Submitted Three questions that keep me up at night as a story · 84 points
HN discussionClaude didn't follow your "Every line must earn its keep. Prefer readability over cleverness. We believe that if carefully designed, 10 lines can have the impact of 1000." from https://github.com/quantbagel/gtinygrad/blob/master/AGENTS.m... given how bloated this demo is. https://blog.evjang.com/2019/11/jaxpt.html is a better demo of how to render the Cornell Box on a TPU using differentiable path tracing.
Comment linking /2019/11/jaxpt.html
on “High speed graphics rendering research with tinygrad/tinyJIT”Submitted Software and Hardware for General Robots as a story · 42 points
HN discussionSubmitted Meta-Learning in 50 Lines of JAX as a story · 35 points
HN discussionSubmitted Dijkstra’s in Disguise as a story · 25 points
HN discussionSubmitted Dijkstra's in Disguise as a story · 9 points
HN discussionSubmitted Eric Jang: Differentiabe Path Tracing on the GPU/TPU as a story · 5 points
HN discussionSubmitted Uncertainty: a Tutorial as a story · 5 points
HN discussionSubmitted What I Cannot Control, I Do Not Understand as a story · 4 points
HN discussionSubmitted NFTs and Financial Nihilism as a story · 3 points
HN discussionSubmitted Science and Engineering for Learning Robots as a story · 3 points
HN discussionSubmitted Robinhood, Leverage, and Lemonade as a story · 3 points
HN discussionSubmitted Meta-Learning in 50 Lines of JAX as a story · 3 points
HN discussionSubmitted How to get an internship as a story · 3 points
HN discussionSubmitted How to Get an Internship as a story · 3 points
HN discussionSubmitted My Internship Experiences at Pixar, Google, and Two Sigma as a story · 3 points
HN discussionSubmitted My Internship Experiences at Pixar, Google, and Two Sigma (2015) as a story · 3 points
HN discussion
