HN user
https://siboehm.com
You can achieve close to the same thing using uBlock element zapper, without installing extra plugins. I use the same trick for SO ('Hot Network Questions' darkpattern) and other sites.
Ha, you're right. I should've used 30,000,000. The answer it gave regarding how to do "ties to even" was still correct though: "Evenness" is decided by the least significant bit of the mantissa, not be the "evenness" of the decimal representation.
Nice, that's simple enough! You may want to try talking to it via the API, last time I checked the web UI doesn't accept prompts >4K tokens while GPT-4 via the API has a 8K Token limit. And then there's the 32K version...
Reading your blogpost it wasn't clear to me how you ran your code through GPT-4. Which prompts did you use?
1. It can do some reformatting tasks faster than I can do them by hand. Example: Inline FuncA into FuncB <paste code for both functions>.
2. For more complicated tasks it requires good prompting. Example: Tell me three ways to fix this error, then pick the best way and implement it. <paste error> <paste relevant code>. Without the "step-by-step" approach it almost never works.
3. It's pretty good at writing microbenchmarks for C++. They always compile, but require some editing. I use the same prompting approach as (2.) for generating microbenchmarks.
4. It's pretty useful for explaining things to me that I then validate later via Google. Example (I had previously tried and failed to Google the answer): The default IEEE rounding mode is called "round to nearest, ties to even". However all large floating point numbers are even. So how is it decided whether 3,000,003 (which is not representable in fp32) becomes 3,000,002 or 3,000,004?.
5. It can explain assembly code. I dump plain objdump -S output into it.
The main limitation seems to be UI. chat.openai.com is horrible for editing large prompts. I wrote some scripts myself to support file-based history, command substitution etc.
Author here: Seems like a good trick! Though won't this affect shared memory alignment and make me loose those LDS.128 instructions? Or do these not require alignment? There's so little good docs on SASS.
In general I'm still confused about whether vectorized load instructions (LDS.128) necessarily lead to bank conflicts or not. My impression was that consecutive 32b floats get mapped to different banks, so to avoid conflicts I'd want the warp to load 32*32b consecutive elements at each step.
I had a brief look at your package, and my impression was that it's only changing model training. If this is correct then the format of the model.txt (calling `lgbm.save(model, "model.txt")`) is the same as regular lightgbm. This would mean you can use my library for inference.
I built this decision tree (LightGBM) compiler last summer: https://github.com/siboehm/lleaves
It get's you ~10x speedups for batch predictions, more if your model is big. It's not complicated, it ended up being <1K lines of Python code. I heard a couple of stories like yours, where people had multi-node spark clusters running LightGBM, and it always amused me because by if you compiled the trees instead you could get rid of the whole cluster.
You can be very worried about the medium-term dangers of AGI even if you believed (which I don't) that consciousness could never arise in a computer system. I think it can be a useful metaphor to compare AGI to nuclear weapons. Currently we're trying to figure out how to make the nuclear bomb not go off spontaneously, and how to steer the rocket. (One big problem w/ the metaphor is that AGI will be very beneficial once we do figure out how to control it, which is harder to argue with nuclear weapons).
Most of these AGI doom-scenarios require no self-awareness at all. AGI is just an insanely powerful tool that we currently wouldn't know how to direct, control or stop if we actually had access to it.
This has always confused me as well. What would be the reason why some adversary would choose to craft an adversarial example and deploy it in the real world versus the much easier solution to just remove / obscure the sign?
Here's a good review of the research around this problem: https://nintil.com/bloom-sigma/
The author concludes (IMO, but you should read it yourself) that while tutoring does have positive effects, the 2 Sigma effect size measured by Bloom was probably an outlier.
I've been running a setup with Recoll and https://github.com/ArchiveBox/ArchiveBox for a few months now [1]. Each morning archivebox scrapes all new links that I've put into my (text-based) notes and saves them as HTML singlefiles. Then Recoll indexes them.
It's very fast and ~4 lines of code. It's surprising how often I rediscover old blog posts & papers that are much better than what Google yields me.
From my experience Recoll isn't very good at searching for aliases sadly.
In stories like this it's important to consider that returns on R&D for these large pharma companies have recently been getting quite low, 1.8% in 2019 according to [0].
[0] https://www.reuters.com/article/us-pharmaceuticals-r-d/pharm...
What helped me the most with sticking to SR and Anki over 6 months now was the tip by Michael Nielsen [0] to not use any plugins and to not personalize anything:
I know many people who try Anki out, and then go down a rabbit hole learning as many features as possible so they can use it “efficiently”. Usually, they're chasing 1% improvements. Often, those people ultimately give up Anki as “too difficult”, which is often a synonym for “I got nervous I wasn't using it perfectly”. This is a pity. As discussed earlier, Anki offers something like a 20-fold improvement over (say) ordinary flashcards. And so they're giving up a 2,000% improvement because they were worried they were missing a few final 5%, 1% and (in many cases) 0.1% improvements. This kind of rabbit hole seems to be especially attractive to programmers.
Emoji might actually be usefull for quick oversight, but how can you include a reference to an issue for every commit? What if you are just adding more tests?