Viewing profile — gaogao
gaogao
HN member- Joined
- Fri, Mar 15, 2019, 4:16 AM UTC
- HN karma
- 860
- Public activity
- 275 items
- HN profile
- View on Hacker News ↗
About gaogao
Recent public activity
-
comment
Comment #49188563
Frequent minor version bumps are pretty common these days. Opus 4.7 -> 4.8 was 42 days.
-
comment
Comment #47175411
Doesn't |& work with bash?
-
comment
Comment #47145810
Nah, the articles are non-contradicting. That article focuses on how the spend mostly goes to imports, which decreases GDP. This one focuses on the effects on unemployment. It's ve…
-
comment
Comment #47098892
Yes, though often the easiest way to replicate it in regular programming languages is to translate that language to Lean or another ITM, though auto-active like Verus is used for R…
-
comment
Comment #47069944
The future is now. The recent couple of Lean releases have leaned more on software verification, I believe in part due to Amazon hiring a number of the core devs.
-
comment
Comment #46933582
Yeah, I pointed a custom thing and Claude at #6, and it's solved it in Lean besides needing to axiomize one theorem not in mathlib. Only about four of the problems have enough foun…
-
comment
Comment #46391640
The compiler for CUDA Tile being Blackwell only is a baffling decision. I wanted to try it out, but it's only really easy to grab H100s quickly right now. I guess maybe I'll try it…
-
comment
Comment #46366996
Yup, I've already spent like $20k using Claude to verify things, so like there's probably some room for cost cutting.
-
comment
Comment #46366961
Yeah, Rust has been pretty good for formal verification so far. Hoare spec contracts I think are the way forward, especially since they fairly naturally flow from unittests. I've b…
-
comment
Comment #46296162
Topical to my interests, I used Claude Code the other day for formally verifying some matrix multiplication in Rust. Writing the spec wasn't too hard actually, done as post-conditi…
-
comment
Comment #46281215
> Season 8: In this season, they introduced gender fluidity. Whilst an interesting problem on its own, this would have wreaked havoc on my model. Well I guess free money except for…
-
comment
Comment #45972567
Pathways, I understand, is more so these days just the name for their training orchestrator for doing distributed JAX stuff - https://github.com/google/pathways-job
-
comment
Comment #45919722
The gaps that led to this was, I think, part of why the CISO got replaced - https://www.thestack.technology/anthropic-new-ciso-claude-cy...
-
comment
Comment #45851903
> In 2007, the scientist, who once worked at the University of Cambridge's Cavendish Laboratory, told the Times newspaper that he was "inherently gloomy about the prospect of Afric…
-
comment
Comment #45839471
If asked non-directly, it still currently answers it - https://www.kimi.com/share/19a5ab4a-e732-8b8b-8000-00008499c...
-
comment
Comment #45825844
Ah so, this is pretty possible, as it's a flavor of text adventure puzzle that usually shows up every year or two for Mystery Hunt. I agree that visually allows this to a greater d…
-
comment
Comment #45681712
Oh also good talk at PTC yesterday! I had meant to ask you more about the formal memory model, but the other post talk questions ended up being really interesting too.
-
comment
Comment #45681694
Yup, hyperreactor, one of the new crates that's part of it, does some particularly interesting things for efficient parallel distributed channels.
-
comment
Comment #45681658
> It's a pity they don't do a complete rewrite with a functional language as the driver. It's open source, so seeing such an extension would be quite cool. There's much that could …
-
comment
Comment #45681619
> As far as things that might be a performance loss here, one thing I'm wondering is if custom kernels are supported Yeah, you might end up needing some changes to remote worker in…
-
comment
Comment #45681580
Nah, focusing on a different controller paradigm. Jax is focused on multi-controller SPMD, while this is focused on a single-controller setup. Both have their place, with single-co…
-
comment
Comment #45681508
Yup, there's stuff like https://pytorch.org/blog/introducing-torchforge/ on top of it now
-
comment
Comment #45675310
Do you think it might be easier to target cuTile instead of PTX? (Probably not, since it has a less formalized model?)
-
comment
Comment #45659376
Yeah, I don't think surrealism or constructed is good in the early data mix, but as part of mid or post-training seems generally reasonable. But also, this is one of those cases wh…
-
comment
Comment #45658984
Brain rot texts seems reasonably harmful, but brain rot videos are often surreal and semantically dense in a way that probably improves performance (such as discussed on this Germa…