Viewing profile — anonymoushn
anonymoushn
HN member- Joined
- Thu, Dec 09, 2010, 7:32 AM UTC
- HN karma
- 6,269
- Public activity
- 3,523 items
- HN profile
- View on Hacker News ↗
About anonymoushn
Recent public activity
-
comment
Comment #48016023
I couldn't reproduce this behavior with Sonnet 4, and Sonnet 3.7 has been deprecated since I messed with this stuff. You can try tokenizing the string " " I think the correct token…
-
comment
Comment #48014208
That's " 'd ".strip(), an english contraction suffix. it's 1 token, but using this echo approach you will be served the apostrophe and the subsequent letter for the first time in d…
-
comment
Comment #47846625
You can't reliably obtain correct token boundaries with this method. For example, "'d" is 1 token, but the API will return "d" stuck to the next token. Weirdly this seems to be spe…
-
comment
Comment #47830680
their old tokenizer performed some space collapsing that allowed them to use the same token id for a word with and without the leading space (in cases where the context usually imp…
-
comment
Comment #47830340
Is this the wrong URL? this seems to be a blog post from October 2025 called "Introducing: Local Browser AI"
-
comment
Comment #47668209
How do you guys decide which settings should be configurable via environment variables but not settings files and which settings should be configurable via settings files but not e…
-
comment
Comment #47668126
> On of our product principles is to avoid changing settings on users' behalf Ideally there wouldn't be silent changes that greatly reduce the utility of the user's session files u…
-
comment
Comment #47602072
Oh, can you post some benchmarks? I didn't know that parser throughput per core would change with the amount of data like that.
- comment
-
comment
Comment #47542039
are those tools known for their fast json parsers?
-
comment
Comment #47052322
ideally users could be banned for posting LLM outputs as if they were authored by humans https://www.pangram.com/history/49335ddf-118d-43e4-9340-a58a...
-
comment
Comment #46655870
Hello, the part about canonical filtering in https://openreview.net/pdf?id=DFybOGeGDS doesn't seem to try to account for pretokenization. For example, if you receive " 天天中彩票APP" in…
-
comment
Comment #46600486
use claude code if you want to use opus
-
comment
Comment #46590619
what does "logprobs look off" mean
-
comment
Comment #46511507
Hello, a couple years ago I participated in a contest to count word frequencies and generate a sorted histogram. There's a cool post about it featuring a video discussing the trick…
-
comment
Comment #46478522
requires fully deterministic inference, which turns out to be unusual, but for this sort of thing it's probably fine if you do really slow inference on cpu. cool idea.
-
comment
Comment #46342342
please write your own posts from now on
-
comment
Comment #46295512
i love stemming, i love searching for "anime" and getting "animal"
-
comment
Comment #46257004
so sad to hear that about Streaming SIMD Extensions
-
comment
Comment #46170481
This is true economically but in reality if you have much larger cost savings than that for sale then these companies mostly say "we would be happy to buy that for $0 while we pay …
-
comment
Comment #46038458
If they had, they would know that it involves many weeks of arguing with support, of course
-
comment
Comment #45965594
The framing in this post is really weird. Automated evals can be much more informative than unit tests because the results can be much more fine grained. A/B testing in production …
-
comment
Comment #45788697
it is frustrating that the post opens by describing latency and then saying that it is called throughput.
-
comment
Comment #45768608
fwiw i thought the previous async based on whole-program analysis and transformation to stackless coroutines was pretty sweet, and similar sorts of features ship in rust and C++ as…
-
comment
Comment #45730129
Arrays of length 0 also exist in json?