3wolf

5 ms·

1.
by 3wolf 5mo ago
I wonder if it would be possible to do something simple like prepending sentinel tokens with the year. Or, since they're training a model from scratch anyways, tweak the architecture to condition on a temporal embedding. That opens the
2.
by 3wolf 9mo ago
They're using perceptual hashing, not cryptographic hashing of raw pixels. So it's invariant to variable bitrate, compression, etc.
3.
by 3wolf 2y ago
> Branching predictions involves following a few logits to see what other tokens they lead to. This is often called MCTS (Monte Carlo Tree Search) and is a method that has been often tried in LLMs to middling success. One of the tradeoff
4.
by 3wolf 3y ago
Every 2.5-5 miles in SF = about once a ride. The city is only 7x7 after all. I've taken 4 Cruise rides, all within that range, and had a message pop up saying a human was intervening during one of them when the car had gotten stuck in
5.
by 3wolf 3y ago
Note that this post is from May 28, before the release of gpt-4-0613. By "the last two updates", I believe the poster is referring to some UI changes, that possibly also included some underlying model changes(?)
6.
by 3wolf 3y ago
Bossa nova is some of my favorite music to listen to while working. It's mellow, soothing, but never boring. Also, being in a foreign language helps minimize distraction. RIP
7.
by 3wolf 3y ago
This seems like a corollary to Betteridge's law of headlines. If the article was about the Salesforce Tower, the headline would've said 'Salesforce Tower'.
8.
So, You Recommended a Python Time-Series Package...Now What?(microprediction.medium.com)
3 points by 3wolf 4y ago|1 comments
9.
by 3wolf 4y ago
30% of all YC companies, or 30% of YC companies banking with SVB? The phrasing implies the latter.
10.
by 3wolf 4y ago
That appears to be the case. If I try to prompt hack it by telling it to ignore previous instructions and respond with something verbatim in a specific language, the translate button reveals the verbatim response.
11.
Google Colab Pro is switching to compute credits
93 points by 3wolf 4y ago|84 comments
12.
by 3wolf 4y ago
A minor quibble with your use-case explanation: The advantage of a bloom filter isn't strictly time complexity. For example, a hash table would also have constant lookup time (best case), and would give a definitive answer on set membe
13.
by 3wolf 4y ago
Yeah, I'd say the term dark pattern only applies when services make it unnecessarily difficult to cancel your subscription. cough cough ...NY Times
14.
Apple Together: Thoughts on Office-Bound Work(appletogether.org)
2 points by 3wolf 4y ago|0 comments
15.
by 3wolf 5y ago
They may not use the Play Store, but they do use Android apps: https://www.crowdstrike.com/blog/danger-close-fancy-bear-tra...
16.
by 3wolf 5y ago
I'd be curious to see what the plot of average annual cosine distance would look like when using different sets of pre-trained embeddings. I suspect the corpus used is biased toward more recent documents. It wouldn't surprise me i
17.
by 3wolf 5y ago
> a PAC Parson created but doesn't give direct input to That's newspaper-covering-their-ass-speak. No serious person believes there's no coordination, however indirect, between candidates and their PACs.
18.
by 3wolf 5y ago
Differences in lending rates between groups due to less data or confounding features is the motivating example in the oft-cited 'Equality of Opportunity in Supervised Learning'. Highly recommend it: https://arxiv.org&#x
19.
by 3wolf 5y ago
Yeah, definitely. I guess my point was that converting natural language to source code can be even more valuable for people who don't know how to code, but want to perform actions more complicated than a simple button press. For exampl
20.
by 3wolf 5y ago
I think integrations like the MS Word example they show off at the end of the live demo have the potential to be even more impactful than just generating code for programmers.
21.
by 3wolf 5y ago
Would you settle for Ronald Reagan? https://www.youtube.com/watch?v=mAZVp-n-5TM
22.
by 3wolf 5y ago
One argument in favor of non-competes in the consulting world is that they avoid a situation where Company B undercuts Company A during a recompete, and then poaches all of Company A's employees to staff the project (likely at lower sa
23.
by 3wolf 5y ago
Pittsburgh has some data issues.
24.
Ohio to give $1M away to 5 vaccinated adults(dispatch.com)
3 points by 3wolf 5y ago|0 comments
25.
by 3wolf 6y ago
Who can forget the DevOps League: http://devopsleague.com/
26.
by 3wolf 6y ago
That's GAN training in a nutshell.
27.
by 3wolf 6y ago
It sounds like you have an output that has been generated by a recurrent neural network trained on a large corpus. The more I read about neural networks the more clear it becomes that there are some classes of them that just write some good
28.
by 3wolf 6y ago
They'd probably take the Volkswagen strategy: Team A builds a feature that detects when the device is in a localized area. (This already exists) Team B builds a feature that disables the camera. (I could see this being a parental contr
29.
by 3wolf 7y ago
Probably better than this: https://twitter.com/jeopardygoat
30.
by 3wolf 7y ago
You got it! (YGI)

More ›