yberreby

6 ms·

31.
by yberreby 1y ago
I'm curious what you think qualifies as science.
32.
Backpropagating through a maze with candle and WASM(yberreby.com)
12 points by yberreby 1y ago|3 comments
33.
by yberreby 1y ago
Better yet, use uv [1]. I've been using it on all of my projects since it came out, and I'm never looking back. It's in a class of its own. [1]: https://docs.astral.sh/uv/
34.
by yberreby 1y ago
The core initial setup here took about 2h30 from getting the laptop out of its packaging to being able to run and develop my main project's code within the environment I described.
35.
by yberreby 1y ago
Interesting that you had such a smooth experience. I was mainly using Homebrew on the daily between 10 and 14 years ago, so I couldn't give you specifics. My experience at the time was poor; maybe I was using it wrong. My impression
36.
by yberreby 1y ago
Care to elaborate on what you've found most painful? Since `nix-darwin` is anything but officially supported, I am expecting trouble, but it would be nice to know if there are specific things I should look out for. I'd love to use
37.
I Used Arch, BTW: macOS, Day 1(yberreby.com)
97 points by yberreby 1y ago|107 comments
38.
by yberreby 2y ago
> Because chatgpt-4o-latest already has all of those improvements Does it, though? They said that "many" have already been incorporated. I simply don't buy their vague statements there. These are different models. They may
39.
by yberreby 2y ago
Good point, but how does that relate to, or explain, the decision not to release 4.1 in ChatGPT? If they have a nice post-training pipeline to make 4o "nicer" to talk to, why not use it to fine-tune the base 4.1 into e.g. chatgpt-
40.
by yberreby 2y ago
I agree on both naming on stability. However, this wasn't my point. They still have a mess of models in ChatGPT for now, and it doesn't look like this is going to get better immediately (even though for GPT-5, they ostensibly want
41.
by yberreby 2y ago
> Note that GPT‑4.1 will only be available via the API. In ChatGPT, many of the improvements in instruction following, coding, and intelligence have been gradually incorporated into the latest version (opens in a new window) of GPT‑4o, a
42.
by yberreby 2y ago
Anecdotally, I used diffrax (and equinox) throughout last year after jumping between a few differential equation solvers in Python, for a project based on Dynamic Field Theory [1]. I only scratched the surface, but so far, it's been a
43.
by yberreby 2y ago
> the latest LLM models have almost as many parameters as we have neurons I often see this take, but the apt comparison is between parameter and synapse count, not neuron count. You should be counting hidden units rather than weights
44.
by yberreby 3y ago
> Actually your comment gives me hope that we will never have AI singularity, since how the brain works is flawed, and were trying to copy it. As someone working in the field, the vast majority of AI research isn't concerned with co
45.
by yberreby 3y ago
> With 5 finetunes you need to host 5 copies or load and unload them. If you use LoRA, which many do when fine-tuning nowadays, you don't need five full copies. You only need to store adapters, which can be in the tens of MBs range
46.
by yberreby 3y ago
Indeed. But a point that is often omitted from comparisons with organic brains is how much "compute equivalent" we spent through evolution. The brain is not a blank slate; it has clear prior structure that is genetically encoded.
47.
by yberreby 3y ago
I'm not denying that there are obvious limitations. However, attributing them to being "pixel-based" seems misguided. First off, the model acts in latent space, not directly on pixels. Secondly, there is no fundamental limi
48.
by yberreby 3y ago
> It just computes next frame based on current one and what it learned before, it's a plausible continuation. ... > Now it's just a "what is the next frame of this 2D image" model plus some textual context. This is
49.
ServerWitch: Grant a custom GPT shell access to your machine in a single command(github.com)
4 points by yberreby 3y ago|0 comments
50.
by yberreby 3y ago
Using LaTeX does not necessarily imply rendering to PDF. https://ar5iv.labs.arxiv.org/
51.
Understanding the Neural Tangent Kernel (2019)(rajatvd.github.io)
2 points by yberreby 3y ago|0 comments
52.
by yberreby 3y ago
I do not mean to discount your experience; I'm simply disproving the assertion that there isn't anywhere normal priced. Montreal is a major city with a lot of tech companies and research activity, so I think it is worth mentioning
53.
by yberreby 3y ago
Hi. I live alone in downtown Montreal. I pay 999 CAD a month for a 1-bedroom apartment, sublet. This includes furniture, electricity, heat, Internet. I have friends who are renting a shared flat, works out to ~650 CAD / person, in a ni
54.
by yberreby 3y ago
I did not know about BGE, learned about it from your comment. Seems to be the new SoTA for semantic embeddings, even for small models. Very cool!
55.
by yberreby 3y ago
As someone who wrote a fair amount of Rust and OCaml code, I have to agree with the author. While working at Routine (YC W21), I was tasked with porting our core library to iOS to minimize duplication of business logic. This was a lucky opp
56.
by yberreby 3y ago
Love the concept. I hope there will eventually be an Emacs version of the extension - I couldn't imagine going back to VS Code.
57.
by yberreby 3y ago
GP didn't say to go for mainstream GC languages, though. OCaml, F#, or Haskell would offer many of the benefits that come with Rust's type system. I've used OCaml professionally, and it's been a real pleasure.
58.
by yberreby 3y ago
Cool project. I tried it last time this got posted, but it was still a bit buggy. Giving it another shot - I'm mainly interested in the local chat. Could you elaborate on the incremental search feature? How did you implement it? Don&#x
59.
by yberreby 3y ago
Indeed, but I'm not claiming it's always faster to check than to produce. Simply that when it is the case , using GPT-4 can be worth it. It is to me - I use it daily.
60.
by yberreby 3y ago
> The problem is distinguishing between these parts requires me to be be an expert in the area I’m inquiring about - and then why the heck do I need to ask some idiot bot for answers to questions that I already know an answer to? Because

More ›