dongping

8 ms·

31.
by dongping 3y ago
In the sense of replicating the results, we do have CI servers and even fuzzers running for our "code replication".
32.
by dongping 3y ago
Then those unconfirmed results are better put on arxiv, instead of being used to evaluate the performance of scientists. Tenure and grant committees should only consider replicated work.
33.
by dongping 3y ago
https://web.archive.org/web/20230130143126/https://blog.ever...
34.
by dongping 3y ago
In some fields, replication is already the prerequisite to benchmark the SoTA. So the incentives boil down to publishing them along with negative results. Or as some have suggested, make it mandatory for PHD candidates to replicate. Though,
35.
by dongping 3y ago
While it is a lot of work, I tend to think that one can then always publish preprints if they can't wait for the replication. I don't understand why a published paper should count as an achievement (against tenure or funding) at a
36.
by dongping 3y ago
"peer code testing" is already the job of the CI server. As it is nothing new, it probably is not going to catch on.
37.
Replace peer review with “peer replication” (2021)(blog.everydayscientist.com)
583 points by dongping 3y ago|337 comments
38.
by dongping 4y ago
It already fails to answer rather simple (but long) multiplication like 975 * 538, even if you tell it do it in a step-by-step manner.
39.
by dongping 4y ago
That's exactly why bitsandbytes has a threshold parameter to control the quantization.
40.
by dongping 4y ago
I have yet to encounter problems with Cuda in our Bazel environment, but I don't have to maintain the Bazel rules, since our build infra team does that. I would think that rules like https://github.com/liuliu/rules
41.
by dongping 4y ago
Honestly, I find the usage of the adjective "modern" deceptive and insincere in the marketing term "Modern CMake". The "modern" scripting language doesn't even support return values and the variables lack
42.
An unexpected rise in IgG4 antibodies after repeated SARS-CoV-2 mRNA vaccination(twitter.com)
4 points by dongping 4y ago|0 comments
43.
by dongping 4y ago
Discord's UI is seemly optimized for dopamine release, just like a game. Some people like this kind of constant distraction, though.
44.
by dongping 5y ago
Bazel would be a better solution in terms of reproducibility and user-friendliness, though.
45.
by dongping 5y ago
China's voting behavior in the security council has been consistently predictable, to the point that Chinese people would laugh at the abstentions at almost every noteworthy issue. Reference: https://jamestown.org/progr
46.
by dongping 5y ago
Zulip is threaded by default, with mandatory topic keywords as title, so pretty much a hybrid chat/forum solution. Maybe you want to take a look at for example https://rust-lang.zulipchat.com/ to see how it works.
47.
by dongping 5y ago
That's the exact reason I find discord annoying. Instant gratification with its feedback means constant distraction. It might be okay for gaming, but using that for collaboration?
48.
by dongping 5y ago
Matrix might be playing the catch-up game, Zulip is IMO better feature-wise for group collaboration because of its threading model, especially for large communities (like Rust). Unless one wants to stick with synchronous voice chatting, I f
49.
by dongping 5y ago
Unpaywalled link: https://archive.md/3uYWZ
50.
What Is the Meaning of Omicron?(forbes.com)
2 points by dongping 5y ago|1 comments
51.
by dongping 5y ago
The totalitarian German government has for example ordered a long light lockdown of the asymptomatic just last year. The maximal time needed for such a lock-down would be something like a month at most 2 months (Wuhan started the lockdown o
52.
by dongping 5y ago
Transmission chains could be broken by quarantine (supported by grocery delivery), masking, simultaneous vaccination and mass PCR testing. It is a coordination problem but it is way cheaper than letting the virus evolve freely. Why do we wa
53.
by dongping 5y ago
Elimination seems to be the only viable strategy to go back to normal, and clearly the cheaper option (after 2 years it is crystal clear). With vaccines and more PCR tests, it would be easier to implement at least for the developed countrie
54.
Elimination vs. mitigation of SARS-CoV-2 in the presence of effective vaccines(thelancet.com)
5 points by dongping 5y ago|0 comments
55.
China Weighs Opening Tencent, ByteDance Content to Search, Sources Say(bloomberg.com)
3 points by dongping 5y ago|0 comments
56.
by dongping 5y ago
The discussion environment is worse because there're more people online. Being online alone was a great filter for fruitful discussions. At least we have Library Genesis and Sci-Hub, which fit the vision of "tech utopia". It
57.
by dongping 5y ago
Individual cost benefit analysis aside, normally vaccinated people would certainly feel safer and goes out more, offsetting the already meager reduction of transmissibility. (Certainly I don't have data to say if it is net benefit or n
58.
by dongping 5y ago
Surely it would be great if you might provide a better description of the event (and the timeline, long after the outbreak in Wuhan).
59.
by dongping 5y ago
The scientific consensus was that there was no evidence that mask would work (or not work) for 2019-nCoV. There's plenty of evidence that it helps prevent transmission of other respiratory viruses. So which hypothesis would have most l
60.
by dongping 5y ago
I think Youtube did a great job NOT banning pro-mask videos which were clearly against CDC recommendation. Or I would suggest that thought police should lock those with the above-mentioned opinions up.

More ›