Lightnews — Scholar-powered news

Shubhendu Trivedi

@shubhendu.bsky.social

830 followers 240 following 3.5K posts

Interests on bsky: ML research, applied math, and general mathematical and engineering miscellany. Also: Uncertainty, symmetry in ML, reliable deployment; applications in LLMs, computational chemistry/physics, and healthcare. https://shubhendu-trivedi.org

shubhendu-trivedi.org

Posts Media Videos Starter Packs

Shubhendu Trivedi @shubhendu.bsky.social · 4h

Good defense talk by Dr. Yoohan (meaning god is gracious).

Johan Edstedt @parskatt.bsky.social · 13h

Yesterday I (successfully) defended my doctoral thesis 🎉
You are now forced to call me Dr. Edstedt.

Public defence of doctoral thesis in Electrical Engineering - Johan Edstedt

Johan Edstedt, at the Department of Electrical Engineering (ISY), defends his doctoral thesis "Towards the Next Generation of 3D Reconstruction”. Opponent…

vimeo.com

Reposted by Shubhendu Trivedi

Maxim Raginsky @mraginsky.bsky.social · 8h

2 2 19

Shubhendu Trivedi @shubhendu.bsky.social · 6h

There is a line in Sátántangó, and also the movie, that goes "They haven’t a clue that it is this idle passivity that leaves them at the mercy of what they fear most." This is basically the central theme of Krasznahorkai's entire oeuvre. Every book is an exploration of a different aspect of it.

Shubhendu Trivedi @shubhendu.bsky.social · 6h

despite their density and imagery, which might suggest otherwise. Sátántangó is about the destruction of a collective will through fear, passivity, and plain ignorance, Werckmeister Harmonies is about the formation of one through the same forces.

1 1

Shubhendu Trivedi @shubhendu.bsky.social · 6h

Never managed to be pretentious enough to power through Tarr in full. Incidentally, while it’s rarely mentioned that the two films/books (Sátántangó and Werckmeister Harmonies, based on Sátántangó and Melancholy of Resistance) are inversions of each other, they are so in very simple ways,

Shubhendu Trivedi @shubhendu.bsky.social · 17h

Don't listen to people who label his stuff comically pretentious etc. They have probably not engaged with any of it!)

1 1

Shubhendu Trivedi @shubhendu.bsky.social · 8h

Generally, I think in a lot of mathematical thinking it is much more productive to think adversarially to start (when presented with statements), building up a coherent mental picture in the process, after which you don't need to.

Shubhendu Trivedi @shubhendu.bsky.social · 8h

its users as adults, because "alignment." But anyway, leaving aside all that. This tendency takes an interesting form in mathematical reasoning. A lot of its mathematical mistakes seem to come from the fact that it operates on your cues and doesn't know how to start thinking against it.

1 1

Shubhendu Trivedi @shubhendu.bsky.social · 8h

Then obviously you must be! But generally, Gemini seems somewhat curt to me. ChatPPT's sycophancy is so annoying that I sometimes can't use it. To get good feedback on something you have to do the adversarial framing yourself, or smuggle it in another way. Claude doesn't believe in treating

Shubhendu Trivedi @shubhendu.bsky.social · 8h

The social ecology around sycophancy can cultivate or amplify dark triad traits, even in people who weren’t strongly predisposed to them. Now it seems like we will have it at industrial scale, and completely frictionless.

Shubhendu Trivedi @shubhendu.bsky.social · 8h

People are not ready for the rate of acceleration of the social reproduction/proliferation of dark triad traits due to LLM sycophancy.

2 2

Shubhendu Trivedi @shubhendu.bsky.social · 12h

Hi Dr. Youhannnn. Congrats!

Shubhendu Trivedi @shubhendu.bsky.social · 17h

Don't listen to people who label his stuff comically pretentious etc. They have probably not engaged with any of it!)

Shubhendu Trivedi @shubhendu.bsky.social · 17h

I had a similar post in 2014, but can't find it. I don't recollect the context for it. The context for the above was around the weird prizes last year (don't usually pay attention to them, but Krasznahorkai really deserved it for more than 10 years.

1 1

Shubhendu Trivedi @shubhendu.bsky.social · 17h

Interesting question, which made me realize I am unable to reduce it to an algorithm. Let me think. :)

Shubhendu Trivedi @shubhendu.bsky.social · 17h

I guess still counts, but last year would have been better.

1 1 5

Shubhendu Trivedi @shubhendu.bsky.social · 1d

PS: Since I way overpost on this website (a bit unusual for me), and because such websites have a lot of literalists -- just joking. I’m poking at the “gradient descent in the brain” line of work, and also because there are a lot of (serious) neuro papers on bloom filter-like neural structures.

Shubhendu Trivedi @shubhendu.bsky.social · 1d

"Does the Brain Implement BM25? Evidence for Lexical Ranking from Intracranial Recordings." Science 799 (7813), 167-171.

Shubhendu Trivedi @shubhendu.bsky.social · 1d

Still haven’t been able to develop the right mental hashing and retrieval frameworks for scanning cs CL efficiently. Perhaps they’re forming up there, waiting to compile and become automatic in another eight months.

Shubhendu Trivedi @shubhendu.bsky.social · Jun 5

Somehow, I don’t like relying on social media for this. I prefer looking through the lists myself, having developed (I think) reasonably good thin-slice judgments of papers that help me sift through them quickly. I can still manage the firehose of cs LG pretty easily otherwise, but not LLM papers.

1 1

Shubhendu Trivedi @shubhendu.bsky.social · 1d

i.e. after a positive reflexive loop. The magical, alchemical, and completely unknown solution is to just prepare enough, and not let the loop happen. :P

Shubhendu Trivedi @shubhendu.bsky.social · 1d

I love talking, and hence tend to get into this reflexive loop (since we are talking Soros and the theory of reflexivity) of taking things for granted, overdoing it (e.g. adding in too many minute details, or too many slides, or too many papers), and then go through a negative reflexive loop.

Shubhendu Trivedi @shubhendu.bsky.social · 1d

Just gave a 1.5 hour talk that I was very happy with, after a string of overly compressed (200 slides in one hour, for example), or hyper, or needlessly detailed, talks. Turns out you need to make sure to just prepare adequately. Who would have thought. :(

1 6

Shubhendu Trivedi @shubhendu.bsky.social · 1d

For AI, something similar will happen eventually. But it will look more violent than it happens to be (because of the capital concentration), and because for a while the narrative will undergo a negative reflexive cycle. But the tech will mature, play out like in the dot com case (but bigger).

Shubhendu Trivedi @shubhendu.bsky.social · 1d

But eventually, it is a whole weird thing where it simultaneously (numerically) larger as a bubble than the dot com era, but also more robust than it long term. The dot com era didn't "finish the internet," the bubble eliminated the excesses for some time and the technology and industry matured.

1 2

Shubhendu Trivedi @shubhendu.bsky.social · 1d

If the spoils of the ZIRP era were more diversified (say into energy, manufacturing, semiconductors), random companies, 5 days old, without even a product plan wouldn't be valued at 20 B. The stock bubble does seem to have some tulip mania aspect to it-based on hysteria and narrative contagion.

1 2

Shubhendu Trivedi @shubhendu.bsky.social · 1d

I think it's not a bubble in the sense of the south sea bubble, or the Dutch tulip mania. Those were actual bubbles--where "actual" means that they were almost fully driven by fraud. In contrast, the bubbly aspect of AI comes from the fact that a huge volume of risk capital is all in one space.

1 14