1/4 Do LLMs understand? "They understand in a way that’s very different from how humans understand," Dileep George, @dileeplearning.bsky.social, of Google DeepMind at the Simons Institute workshop on The Future of Language Models and Transformers. Video: simons.berkeley.edu/talks/dileep...

1/2 "LLMs are amazing, but hard to control. Diffusion models are also amazing, and (maybe too) easy to control." @cornelluniversity.bsky.social's Kilian Weinberger, at the Simons Institute, on how these contrasting properties can be used to steer LLMs. Video: simons.berkeley.edu/talks/kilian...

1/4 LLMs learn language competency, factual common knowledge and factual tail knowledge. Factual knowledge has a heavy tail and should be stored in a database. “It doesn’t belong in the model weights”: Cornell's Kilian Weinberger at the Simons Institute. Video: simons.berkeley.edu/talks/kilian...

1/2 "In the LLM scenario...jailbreaking is very powerful. In agentic scenarios, it gets even more dangerous." @sivareddyg.bsky.social of @mila-quebec.bsky.social at the Simons Institute's workshop on Safety-Guaranteed LLMs Video: simons.berkeley.edu/talks/siva-r...

1/2 "You could have bad goals and you could be smart, but if you can’t do anything in the world, then you can’t do a lot of harm. The trio is the thing that kills us." @yoshuabengio.bsky.social, at his Richard M. Karp Distinguished Lecture at the Simons Institute: www.youtube.com/watch?v=g0lj...

1/2 "Symmetries play a fundamental role in machine learning." Soledad Villar of @jhu.edu speaking at the Simons Institute workshop on Randomness, Invariants, and Complexity. Video: simons.berkeley.edu/talks/soleda...

Join us next Tuesday, 12/9 for Jon Kleinberg's talk in our Theoretically Speaking public lecture series. Registration is required. simons.berkeley.edu/events/ais-m...

1/3 "What are the conditions for an AI system in the future to cause catastrophic harm?" Turing Award winner @yoshuabengio.bsky.social asked, during his Richard M. Karp Distinguished Lecture at the Simons Institute earlier this year. www.youtube.com/watch?v=g0lj...

1/2 Networks are creating bottlenecks for AI. There's "a separation between the growth of computation for [AI] and the growth of networking infrastructure...We must make our networks more efficient." Chen Avin of @bengurionuni.bsky.social at the Simons Institute. simons.berkeley.edu/talks/chen-a...

Join us for Jon Kleinberg's Theoretically Speaking lecture, 12/9! This is an in-person event, and registration is required. simons.berkeley.edu/events/ais-m...

1/2 "What are graph foundation models? How are they different from language models? Or maybe not," asked Mikhail Galkin of @googleresearch.bsky.social at the Simons Institute's workshop on Graph Learning Meets Theoretical Computer Science. Video: simons.berkeley.edu/talks/mikhai...

We were delighted to host Simons Institute Industry Day 2025 today, full of lively exchanges and talks by partner companies, program organizers, and postdoctoral-level research fellows.

1/2 "In control and automation, the big change is the advent of foundation models and the potential use of those in automating systems...The challenge is of course for safety." Claire Tomlin of @ucberkeleyofficial.bsky.social on AI for Safety Critical Control. simons.berkeley.edu/talks/claire...

1/2 "It seems silly, but it's a very important problem." Virginia Vassilevska Williams (@mit.edu) on the progress in matrix multiplication algorithms during her Richard M. Karp Distinguished Lecture On Matrix Multiplication Algorithms at the Simons Institute. simons.berkeley.edu/talks/virgin...

Ever wondered about graph learning? Watch Ameya Velingker and Haggai Maron give a masterful introduction at the Simons Institute's workshop on Graph Learning Meets Theoretical Computer Science. Video: simons.berkeley.edu/talks/ameya-...

We’re delighted to be supported by the Google DeepMind x Google.org AI for Math Initiative, which was launched today. The new generation of AI tools will transform research on the foundations of computing, and this new initiative is poised to accelerate that. blog.google/technology/g...

Today at 3:30 p.m. PT. Join us! simons.berkeley.edu/events/matri...

Join us next Tuesday! simons.berkeley.edu/events/matri...

1/2 Should have paid attention to matrices during linear algebra classes! In 2026, AI will use ~1% of global electricity, of which ~45-90% will be for matrix multiplications, said Oded Schwartz of Hebrew University of Jerusalem at the Simons Institute. simons.berkeley.edu/talks/oded-s...

Join us for Robert Tarjan's Richard M. Karp Distinguished Lecture on "Asynchronous Concurrency in Data Structures," Tuesday, October 21 at 3:30 p.m. Pacific Time. simons.berkeley.edu/events/async...

1/2 Matrix multiplications are central to machine learning. UC Berkeley Professor Olga Holtz's analyzed, from scratch, the arithmetic complexity of matrix multiplication using Strassen's fast algorithm. She spoke at the Simons Institute. Video: simons.berkeley.edu/talks/olga-h...

Join us! simons.berkeley.edu/events/async...

It was a pleasure chatting with Yael and Daniele about several developments in #cryptography! Thank you @simonsinstitute.bsky.social for this opportunity!

Simons Institute Senior Scientist Nikhil Srivastava on Diagonalization Algorithms, from the Complexity and Linear Algebra Boot Camp. simons.berkeley.edu/news/diagona...

We are currently welcoming applications for postdoctoral-level research fellowships in our Fall 2026 programs. Apply by November 15! simons.berkeley.edu/research-fel...

Quantum Industry Day 2025 Thursday, Oct. 16, 2025 9:30 a.m. – 5 p.m. PT Register for the livestream and to apply to attend in person. Space is limited. We will contact you if we're able to offer you an in-person spot. simons.berkeley.edu/events/quant...

In this episode of Polylogues, @lakshmic.bsky.social sits down Yael Tauman Kalai and Daniele Micciancio to discuss developments in lattices and FHE, the gap between theory and practice, and the unique culture of crypto as a field. www.youtube.com/watch?v=n_R3...

As part of the Algorithmic Foundations for Emerging Computing Technologies Boot Camp, David Patterson reviewed the drivers of computer architecture and upcoming critical challenges. www.youtube.com/watch?v=EvlZ...

Party time

Spring 2025 Simons Institute Law and Society Fellow Serena Booth (Brown University) was an AI policy advisor in the US Senate before joining us. Read her feature in our October newsletter, "Rebuilding an Optimistic Vision for AI Policy." simons.berkeley.edu/news/rebuild...

1/2 Focusing on the 4Ms of AI—model (FLOPs), machine (GPU, mWh / 1.0e12 FLOPs), mechanization (power usage effectiveness) and maps (sites w/ clean energy)—can reduce AI’s carbon footprint, said UC Berkeley's David Patterson, at the Simons Institute. Video: simons.berkeley.edu/talks/david-...

Congratulations to Thomas Rothvoss and Lang Liu, the inaugural winners of the Trevisan Prize at Bocconi University. cs.unibocconi.eu/trevisan-pri...

1/2 "Bad AI carbon footprint commandments." @ucberkeleyofficial.bsky.social's David Patterson's "don'ts" for reducing AI's carbon footprint: don't pick the biggest model; don't pick an older GPU; don't pick a local data center. He spoke at the Simons Institute. simons.berkeley.edu/talks/david-...

Applications are open for research fellows and senior visitors for the Fall 2026 programs on Spectral Theory Beyond Graphs, and on Pseudorandomness and High-Dimensional Expansion. simons.berkeley.edu/research-fel... simons.berkeley.edu/long-term-vi...

Join us for Mark Embree's Richard M. Karp Distinguished Lecture next Tuesday! Register to attend in person or view the livestream. simons.berkeley.edu/events/compu...

Next week at the Simons Institute, a workshop on Linear Systems and Eigenvalue Problems. Join us! simons.berkeley.edu/workshops/li...

From our friends at CLIMB: Voleon Seminar with Ariel Procaccia "Thinking Outside the Ballot Box" Friday October 3rd, 2 p.m. PT 306 Soda Hall (HP Auditorium) Details: climb.berkeley.edu/thinking-out... Livestream: www.youtube.com/channel/UC9t...

1/2 How can one make sense of dolphin vocalizations and behavior? First, "use machine learning to rediscover human words by just watching unlabeled videos of people talking about things," said @mit.edu's Mark Hamilton, speaking at the Simons Institute. Video: simons.berkeley.edu/talks/mark-h...

Join us Tuesday at 3:30 p.m. PT for the first Karp Lecture of the year. Register to attend in person or view the livestream. simons.berkeley.edu/events/virtu...

1/2 While in Hawaii during the pandemic, Markus Freitag realized he should take whale language translation seriously. "I really fell in love with whales. You could just lay in the water and you could hear them singing," he said at the Simons Institute. Video: simons.berkeley.edu/talks/markus...

Here's my latest blog alert on sublinear #algorithms and #differentialprivacy in a wide-ranging conversation with @ccanonne.github.io! scieye.wordpress.com/2025/09/17/e... #randomness #BigData #SciComm @simonsinstitute.bsky.social @sydney.edu.au

First was a whole slate of talks from the @simonsinstitute.bsky.social Algorithmic Foundations for Emerging Computing Technologies symposium: Kaushik Roy - neuro-inspired algorithms and hardware www.youtube.com/watch?v=KvTd... (2/6)

Next week at the Simons Institute, a workshop on Managing Memory. Hope to see you there! simons.berkeley.edu/workshops/ma...

Register to attend, view the livestream, or access the videos before they're captioned for YouTube. simons.berkeley.edu/events/virtu...

Join us next Tuesday! simons.berkeley.edu/events/virtu...

Adam Kalai (OpenAI) explores how to evaluate machine translation systems in the absence of ground-truth reference translations, focusing on the extreme case where only acoustic outputs are available, without contextual or visual grounding. simons.berkeley.edu/news/validat...

In his talk in our workshop with @ProjectCETI on Decoding Communication in Nonhuman Species, Markus Freitag surveyed the rise of LLM-driven translation and its near-human performance in high-resource languages. simons.berkeley.edu/news/machine...

1/2 When the Beatles sang Octopus’s Garden, they could not have imagined real-life Coral Gardeners working with “the coral that lies beneath the waves.” Rich Williams spoke of using AI to restore coral reefs at scale at the Simons Institute's. Video: simons.berkeley.edu/talks/rich-w...

It's not too late to catch up on the talks from the Complexity and Linear Algebra Boot Camp, happening this week. Register to attend in person, watch the live stream, or access the talk videos before they're captioned for YouTube. simons.berkeley.edu/workshops/co...

Join us for the first Richard M. Karp Distinguished Lecture of 2025–26! simons.berkeley.edu/events/virtu...