sdenton4

30 ms·

691.
by sdenton4 3y ago
In the sense of existence, sure. But it's pretty clear that one side has gone far, far further down the conspiracy hole. Can you think of anything in the dem side as widespread as, say, the ivermectin+antivax crap was? This stuff isn&#
692.
by sdenton4 3y ago
Consider the general prevalence of conspiracy theories on the American right - Alex Jones, qanon, stop the steal, etc. Consider that many on the left avoid jobs in the military, where most of the UFO stuff is coming from. I'm on team &
693.
by sdenton4 3y ago
I expect the answer is some combination of economy of scale (they sell lots more cars than walls) and subsidies....
694.
by sdenton4 3y ago
I mean, the simplest answer is a matmul... Given embedding x, y, find M such that Mx ~= y. Easy to train so long as you've got access to both models to compute embedding over whatever you're interested in... (easy to extend to two
695.
by sdenton4 3y ago
The famous reform mayor of Bogota, Antanas Mockus, also started in math! https://en.wikipedia.org/wiki/Antanas_Mockus
696.
by sdenton4 3y ago
Yes, my math grad student time ws an excellent preparation for coding interviews. In addition to the oral exams, I found TA'ing was also an excellent preparation. Students show up with lots of questions, which you get to solve on your
697.
by sdenton4 3y ago
All the ox tongues hanging out of people's bags, remnants of drool and blood pooling on the floor...
698.
by sdenton4 3y ago
They were fine for the first third to half of the game, but eventually the thin quest engine starts showing pretty badly...
699.
by sdenton4 3y ago
Your problem is harder than it needs to be; we can convert it to a closed problem, by requiring citations. Then the question is whether the statement is properly supported by the citations, which is far easier to evaluate.
700.
by sdenton4 3y ago
It's a very promising idea, though. Judging content is much easier than creating content, after all.
701.
by sdenton4 3y ago
Ha. Afaict, most triple-a open world games are built on pretty simple engines, and the great majority of quests fit in a relatively small range of formats. Go to town x, find NPC y, go to landmark z, fight some monsters, get the macguffin,
702.
by sdenton4 3y ago
This is laughable... Neural networks also have basically unknowable abstractions encoded in their weights. There was some work not long ago which taught an ANN to do modular arithmetic, and found that it was doing Fourier transforms with th
703.
by sdenton4 3y ago
Nah taxes solve a variety of collective action and common goods problems that we are apparently unable to solve otherwise. They are to be celebrated, not simply as a necessary evil.
704.
by sdenton4 3y ago
We just need a Netflix series about a geometry prodigy with excellent fashion sense.
705.
by sdenton4 3y ago
For euclidean distance, we have (q-v)^2 = qq + vv - 2qv. The qq and vv don't tell us anything about the relationship between q and v, so are irrelevant to search. So really we are mostly trying to maximize inner product when we search.
706.
by sdenton4 3y ago
Specialists who carefully wield their HN up votes to make sure you see stories like this one...
707.
by sdenton4 3y ago
The pointing finger is not the moon... 1 is the mathematically correct answer, and the floating point behavior is a noisy approximation.
708.
by sdenton4 3y ago
Are you suggesting we should move on from drone wars to clone wars? https://youtu.be/enrUBmDgeOQ?si=F3nfNjO40_v04wCu&t=17
709.
by sdenton4 3y ago
Yeah, it really comes down to the question of how we advance on just-bits vs constrained-environment robotics vs open-domain robotics... Some interesting work here on using LLMs to improve on open-domain robotics: https://arstech
710.
by sdenton4 3y ago
For my group, there's journals and a whole ecosystem outside of the ml conference space related to the actual application area... We can go to a top-tier journal in the application domain and get better reviewers.
711.
by sdenton4 3y ago
+1 to this. I've been orienting away from ML conferences because the review process is so bad; it's an incredible amount of wasted work responding to and re-submitting things after getting a bad crop of reviewers. Honestly it also
712.
by sdenton4 3y ago
Note that 'no free lunch' has a specific meaning with no relation whatsoever to model size/quality trade-offs... https://en.wikipedia.org/wiki/No_free_lunch_theorem In the speed/quality trade-off se
713.
by sdenton4 3y ago
Bitcoin specifically burns electricity like mad... We cry because, in an effort to demonstrate utility of the workload, Bitcoin would burn resources simply to enable transactions. It's completely unnecessary overhead: Ethereum's p
714.
by sdenton4 3y ago
If you have to pay $X for AI inference, then it proves that the work is worth at least $X to you... Which suggests the obvious marketplace scheme: Compute providers are paid (or credited) $X/k to contribute their nodes to the network,
715.
by sdenton4 3y ago
Start spamming senators with dmca takedowns for funding pitch videos?
716.
by sdenton4 3y ago
Lovely paper, thanks!
717.
by sdenton4 3y ago
The thing is, we typically have three parts of a compression scheme: the data, the model (dictionary, Huffman tree, etc), and the program. The model is typically learned from the data at compression time. The program itself is general, howe
718.
by sdenton4 3y ago
So has anyone tried lossless compression with an LLM? Method: create a list of next token predictions, ordered by describing probability, and encode the actual token as its position in the ordered list (so, eg, the most likely token is enco
719.
by sdenton4 3y ago
If the set theorists knew how to farm, then we wouldn't need the NSF to give them money.
720.
by sdenton4 3y ago
They're different kinds of easy... Writing fresh code is easy because you just express your ideas. What's hard is hitting all the contingencies appropriately; it can take years of 'burn in' to get right. Reading code is

More ›