sanxiyn

12 ms·

91.
by sanxiyn 1y ago
FYI: Direct Multipixel Imaging and Spectroscopy of an Exoplanet with a Solar Gravity Lens Mission. https://arxiv.org/abs/2002.11871
92.
by sanxiyn 1y ago
Yes I have.
93.
by sanxiyn 1y ago
No? With structured outputs you get valid JSON 100% of the time. This is a non-problem now. (If you understand how it works, it really can't be otherwise.) https://openai.com/index/introducing-structured-outputs-in
94.
by sanxiyn 1y ago
As far as I know there is no technical reason why jemalloc shouldn't be the default allocator. In fact, as pointed out in the article, it IS the default allocator on FreeBSD. My understanding is it is largely political.
95.
by sanxiyn 1y ago
mimalloc is a good choice. CPython recently switched to mimalloc.
96.
by sanxiyn 1y ago
Interestingly, as written, it is ΚΑΚΙωΝ not ΚΑΚΙΩΝ.
97.
by sanxiyn 1y ago
I don't live in USA lol
98.
by sanxiyn 1y ago
I think it is exactly this. There is no doubt whatsoever OpenAI can do this, but they decided not to. The reason, I think, is that they don't want to be a couple of months late. In other words, they spent $3B to save a couple of months
99.
by sanxiyn 2y ago
Possibly. I have seen Claude switching to Russian for a word or two when it is about revolution!
100.
by sanxiyn 2y ago
Sonnet 3.7's 70% is without thinking, see https://www.anthropic.com/news/claude-3-7-sonnet
101.
by sanxiyn 2y ago
The average human score on USAMO (let alone IMO) is zero, of course. Source: I won medals at Korean Mathematical Olympiad.
102.
by sanxiyn 2y ago
AIME is so not IMO.
103.
by sanxiyn 2y ago
Nope, no LLMs reported 50~60% performance on IMO, and SOTA LLMs scoring 5% on USAMO is expected. For 50~60% performance on IMO, you are thinking of AlphaProof, but AlphaProof is not a LLM. We don't have the full paper yet, but clearly
104.
by sanxiyn 2y ago
As documented in http://cs.bell-labs.co/who/dmr/chist.html , the primary incompatibility is uses of =+ for +=. Grep for =+ in the repository.
105.
by sanxiyn 2y ago
> What's the A+ example out there of a unionized engineering team that has been able to find a great work-life balance, great benefits, and a fun product development life-cycle that is profitable or clearly on its way to profitabili
106.
by sanxiyn 2y ago
I used to think only proposal for quantum supremacy on near future NISQ devices is random circuit sampling, for which I think "generate random numbers" is a fair summary. Then someone told me about NISQ-TDA and I can't immedi
107.
by sanxiyn 2y ago
Scaleway is pretty competitive.
108.
by sanxiyn 2y ago
Because US proved itself to be an unreliable partner.
109.
by sanxiyn 2y ago
Go Scaleway! I always wanted to root for it.
110.
by sanxiyn 2y ago
Proton seems to be there?
111.
by sanxiyn 2y ago
Since it is a library, it won't be named lynx anyway even if there is no name collision. D3 is packaged as libjs-d3, for example.
112.
by sanxiyn 2y ago
So far, I don't think we found anything interesting, yet.
113.
by sanxiyn 2y ago
From the blog post: > We plan to provide additional details in the upcoming months on how we reverse engineered the microcode update process, which led to us identifying the validation algorithms
114.
by sanxiyn 2y ago
zlib-rs does use ARM CRC32 instructions though.
115.
by sanxiyn 2y ago
Everyone is a typist now, so I don't think it is farfetched that everyone is a SWE in the future.
116.
by sanxiyn 2y ago
If so kernel is released with broken Rust. That is the policy, and I am flabbergasted why everyone is going "that policy must not be literal".
117.
by sanxiyn 2y ago
Are you familiar with Linux kernel development process? Features can be merged only in two weeks long merge window. After the merge window closes, only fixes are merged for eight weeks. Rust binding can be fixed in that time. I don't s
118.
by sanxiyn 2y ago
Emotionally I agree, but... o1 was a paradigm shift. Nothing DeepSeek has done is on that level yet. DeepSeek themselves would agree. Supposedly Liang Wenfeng himself flew to US to gather information when o1 was launched.
119.
by sanxiyn 2y ago
Why wouldn't it be merged? No Rust code is built unless CONFIG_RUST is on, and it is off by default. It won't be on by default for a long time.
120.
by sanxiyn 2y ago
Many people replied with anecdotes, but recently Anthropic published analysis of claude.ai production traffic. As far as I can tell this is the single best currently existing source of "how people actually use LLMs". Everything el

More ›