@lschiff@mastodon.sdf.org2026-05-26 16:23:45
Government #censorship is having a direct impact on federally funded research, restricting the topics addressed and the experiences & perspectives recognized. Join Jeremy Berg, Former #NIH institute director and Editor-in-Chief, Science magazine, Michael D. Green, Ph.D., Early Career Researcher…
@NuclearDisorder@mastodon.social2026-05-26 06:05:07
Heute vor 68 Jahren: Am 26.05.1958 kam es bei Enewetak zum Atomtest Operation Hardtack I, "Magnolia". Dieser Test war Teil einer Serie von 35 #Atomtests, die die USA im Sommer 1958 auf den #Marshallinseln im Pazifik durchführten.
@arXiv_physicschemph_bot@mastoxiv.page2026-03-27 08:44:42
Enabling ab initio geometry optimization of strongly correlated systems with transferable deep quantum Monte Carlo
P. Bern\'at Szab\'o, Zeno Sch\"atzle, Frank No\'e
https://arxiv.org/abs/2603.25381 https://arxiv.org/pdf/2603.25381 https://arxiv.org/html/2603.25381
arXiv:2603.25381v1 Announce Type: new
Abstract: A faithful description of chemical processes requires exploring extended regions of the molecular potential energy surface (PES), which remains challenging for strongly correlated systems. Transferable deep-learning variational Monte Carlo (VMC) offers a promising route by efficiently solving the electronic Schr\"odinger equation jointly across molecular geometries at consistently high accuracy, yet its stochastic nature renders direct exploration of molecular configuration space nontrivial. Here, we present a framework for highly accurate ab initio exploration of PESs that combines transferable deep-learning VMC with a cost-effective estimation of energies, forces, and Hessians. By continuously sampling nuclear configurations during VMC optimization of electronic wave functions, we obtain transferable descriptions that achieve zero-shot chemical accuracy within chemically relevant distributions of molecular geometries. Throughout the subsequent characterization of molecular configuration space, the PES is evaluated only sparsely, with local approximations constructed by estimating VMC energies and forces at sampled geometries and aggregating the resulting noisy data using Gaussian process regression. Our method enables accurate and efficient exploration of complex PES landscapes, including structure relaxation, transition-state searches, and minimum-energy pathways, for both ground and excited states. This opens the door to studying bond breaking, formation, and large structural rearrangements in systems with pronounced multi-reference character.
toXiv_bot_toot
@dawid@social.craftknight.com2026-05-23 07:08:00
Wczoraj włochy vs dziś niemcy - wczoraj przydrożny bar w dolinie aosty - zamówienie całe 7€. Dzisiaj super piekarnia w małym bawarskim miasteczku - 10,50€ za całość...
Ehh, byle do pierwszego paczkomatu, żeby odebrać kawę z naszej ulubionej palarnii i już jeść w domu -.-
#microblog
@NuclearDisorder@mastodon.social2026-05-26 06:04:58
Heute vor 68 Jahren: Am 26.05.1958 kam es bei Enewetak zum Atomtest Operation Hardtack I, "Yellowwood". Dieser Test war Teil einer Serie von 35 #Atomtests, die die USA im Sommer 1958 auf den #Marshallinseln im Pazifik durchführten.
@Techmeme@techhub.social2026-04-08 15:10:56
Alibaba and China Telecom launch a data center in southern China that is powered by 10,000 of Alibaba's Zhenwu chips designed for AI training and inferencing (Arjun Kharpal/CNBC)
https://www.cnbc.com/2026/04/08/china-alibaba-data-center-ai-chips-zhenwu.html…
@heiseonline@social.heise.de2026-06-12 05:02:00
@UrbanNature@nerdculture.de2026-04-15 19:19:15
Tipp: Die Stiftung Naturschutz Berlin bietet am 23.4.26 von 16-18 Uhr einen kostenlosen Online-Vortrag zum Thema Libellen an 🧚♂️
Hier anmelden:
https://www.umweltkalender-berlin.de/angebote/details/89412?dat=2026-04-23?dat=2026-04-23…
@aredridel@kolektiva.social2026-04-14 14:22:42
So to follow up on this, I've caught it in action. Models, when quantized a bit, just do a bit more poorly with short contexts. Even going from f32 (as trained) to bf16 (as usually run) to q8 tends to do okay for "normal" context windows. And q4 you start feeling like "this model is a little stupid and gets stuck sometimes” (it is! It's just that it's still mostly careening about in the space of "plausible" most of the time. Not good guesswork, but still in the zone). With long contexts, the probability of parameters collapsing to zero are higher, so the more context the more likelihood you are to see brokenness.
And then at Q2 (2 bits per parameter) or Q1, the model falls apart completely. Parameters collapse to zero easily. You start seeing "all work and no play makes jack a dull boy” sorts of behavior, with intense and unscrutinized repetition, followed by a hard stop when it just stops working.
And quantization is a parameter that a model vendor can turn relatively easily. (they have to regenerate the model from the base with more quantization, but it's a data transformation on the order of running a terabyte through a straightforward and fast process, not like training).
If you have 1000 customers and enough equipment to handle the requests of 700, going from bf16 to q8 is a no-brainer. Suddenly you can handle the load and have a little spare capacity. They get worse results, probably pay the same per token (or they're on a subscription that hides the cost anyway so you are even freer to make trade-offs. There's a reason that subscription products are kinda poorly described.)
It's also possible for them to vary this across a day: use models during quieter periods? Maybe you get an instance running a bf16 quantization. If you use it during a high use period? You get a Q4 model.
Or intelligent routing is possible. No idea if anyone is doing this, but if they monitor what you send a bit, and you generally shoot for an expensive model for simple requests? They could totally substitute a highly quantized version of the model to answer the question.
There are •so many tricks• that can be pulled here. Some of them very reasonable to make, some of them treading into outright misleading or fraudulent, and it's weirdly hard to draw the line between them.
@inthehands@hachyderm.io2026-05-18 19:05:50
RE: https://mastodon.social/@heidilifeldman/116596292735099594
There will surely be — must already be! — many other such efforts sneaking under the radar, exploiting the marginalized, or cake-walking past the indifferent.
The thing about having trillions of investment dollars behind your industry is that you don’t really even need a marketing •strategy• per se; you just need to flood the zone, flood the containers of society and see where the leaks are. Efforts like the one in the linked @… article only need to succeed 1 time in 100 to pay off, because once you’ve got the data you can milk it forever.