abhgh

7 ms·

31.
by abhgh 1y ago
Active Learning is a very tricky area to get right ... over the years I have had mixed luck with text classification, to the point that my colleague and I decided to perform a thorough empirical study [1], that normalized various experiment
32.
by abhgh 1y ago
Evals somehow seem to be very very underrated, which is concerning in a world where we are moving towards (or trying to) systems with more autonomy. Your skepticism of "llm-as-a-judge" setups is spot on. If your LLM can make mista
33.
by abhgh 1y ago
I noticed the conflation of terms too but it seems to arise out of the original Microsoft announcement! https://devblogs.microsoft.com/commandline/edit-is-now-open-...
34.
by abhgh 2y ago
You're right! My bad. Thank you for pointing it out! Leaving my comment up for info on bayesopt.
35.
by abhgh 2y ago
Note the website (ai-contest.com) that the post links to seems to have been hijacked by a gambling site. For the use-cases where Genetic Programming was popular, I would recommend looking at Bayesian Optimization (bayesopt) as an alternativ
36.
by abhgh 2y ago
The people who founded Ponoc seemed to have creative differences with Miyazaki. They wanted to make a movie [1] that they felt Ghibli won't greenlight [2] - but there seems to have been no deep seated animosity or desire to rip-off. In
37.
by abhgh 2y ago
Same here, and also agree with the "prime" notation, i.e., y'! One of my instructors had joked that the dot/prime notations saved printing costs in the days when printing was expensive - not sure if there is any truth to
38.
by abhgh 2y ago
I highly recommend the course you've mentioned (by Yaser Abu-Mostafa). In fact I still recommend it for picking up the basics; very good mix of math and intuition, Abu-Mostafa himself is a terrific teacher, and he is considerate and th
39.
by abhgh 2y ago
I came across this a few days ago, and my excuse to give it a a serious look is that Andreas Krause has some deep and interesting research in Gaussian Processes and Bandits [1]. [1] https://scholar.google.com/scholar?start=1
40.
by abhgh 2y ago
Oh thats a cool fact to know! Thanks.
41.
by abhgh 2y ago
I have mentioned this elsewhere online but I was once attending a talk in the computer history museum for which I had turned up early. An elderly gentleman took his seat right before me - it took me a while to process that it was Knuth (in
42.
Evaluating LLMs – Notes on a NeurIPS'24 Tutorial(blog.quipu-strands.com)
1 points by abhgh 2y ago|0 comments
43.
by abhgh 2y ago
It's not just the width of the column - there are annotations on certain lines (that appear on a right "margin") that don't show up on mobile. I think that makes it not an easy fix, but to your larger point, this is not
44.
by abhgh 2y ago
YouTube recommendations have worked very well for me. I listen to a lot of talks and if it weren't for the recommendations I wouldn't have discovered some very interesting talks or even areas. The only time the recommendations der
45.
by abhgh 2y ago
I have had some success with PowerPoint animations. Put up the equation, and for each mouse click surround the symbol(s) you want to explain in a red box or something and have a infobox show up (which can also contain images, such as plots
46.
by abhgh 2y ago
It is pleasantly surprising to see how close your pipeline is to mine. Essentially a good representation layer - usually based on BERT - like minilm or MPNet, followed by a calibrated linear SVM. Sometimes I replace the SVM with LightGBM if
47.
by abhgh 2y ago
Do these Indonesian gambling sites somehow exploit the AMP cache? The apex domain of my blog was recently hijacked by such a site. I was only using the blog.* subdomain. One day I noticed on the Google search console that someone had verif
48.
by abhgh 2y ago
Reg. #2 - I returned to frontend dev. after years for a hobby project (and even earlier, my frontends weren't very complex, I could get by with jQuery in the pre-React days), and I am surprised to see this is popular (or at least a sup
49.
by abhgh 2y ago
Could you point to some good resources for either, that enables quick application?
50.
by abhgh 2y ago
I second this as a good starting point. When I was picking up z3, this was helpful.
51.
by abhgh 2y ago
I strongly advocate this! If you're starting off in this space, check if this barebones implementation isn't all you need. You can't beat the accuracy of a full search; in theory, you're trading off scalability, but v
52.
by abhgh 2y ago
This was my first big purchase after I finished my Masters. Absolutely loved it. I would often pick a random volume out of the collection, open it to a random page and read. Delightful books, good memories.
53.
by abhgh 2y ago
I feel like we are debating slightly different perspectives, and with that lens I agree with what you say. Here is the difference in perspectives (and this is decoupled from this particular paper): your take is that today the reviews work i
54.
by abhgh 2y ago
Thanks, yes, lot of good ideas in ML seem to be slowly vanishing from the collective awareness. I have nothing against the current spate of methodologies which are empirically great - and if one needs proof, I am a "happy customer&quot
55.
by abhgh 2y ago
Thank you!
56.
by abhgh 2y ago
Thank you for these comments. I appreciate them and I'll consider them in my next draft. However, I would like to point out a few things; just so that we have the larger picture in mind. Again, I do appreciate you took the time to look
57.
by abhgh 2y ago
In my view, in most cases ACs now perform one of two roles (1) summarize what reviewers have said, and/or (2) if there is a high variance in scores, they request the reviewers to sort it out among themselves, and THEN summarize :). I a
58.
by abhgh 2y ago
Yes, it did come up during my defense, but it was deemed not to be a concern since I had one prior paper [1] (the original one in this thread of work, the paper I linked above was an improvement over it), and my advisor (co-author on both p
59.
by abhgh 2y ago
Yes, that's the one: https://arxiv.org/pdf/1906.06852
60.
by abhgh 2y ago
Personal sad story, but hopefully relevant: during my recent PhD I worked on a problem where I used a Dirichlet Process in my solution. That paper has been bouncing around for the past few years getting rejected from every venue I have subm

More ›