@alphaxiv.org released quickarXiv Swap arxiv → quickarxiv on any paper URL to get an instant blog with figures, insights, and explanations.

Interestingly, it's more about some notable high-quality outlets underperforming on engagement, rather than low-quality outlets overperforming:

This morning’s view + final fantasy soundtrack? *chef’s kiss*

My sister has introduced her childhood favorite sailor moon to the nieces. And all I can say is major solidarity with Sailor Pluto today.

illinois!durbin is 202-224-2152; "experiencing high call volumes" but keep trying and in the meantime send an email pronto www.durbin.senate.gov

My partner has started reading Wendell Berry. And look, nothing against Wendell Berry, but is it also bringing back memories of insufferably preachy acolytes in grad school? Yes. My Sunday morning is now spent reflecting on my visceral reactions to didactic agrarians.

Digital humanities researchers often care about fine-grained similarity based on narrative elements like plot or tone, which don’t necessarily correlate with surface-level textual features. Can embedding models capture this? We study this in the context of fanfiction!

I’m a machine that transforms free Sunday afternoons into walks along the arboretum waterfront trail.

you are not going to believe it, but pringles may not even be the best gem in this paper.

Absolute banger

I’ve been in a Halloween costume rut, but the gay Halloween meme format is truly inspirational.

Kpop demon hunters and baseball are doing me a solid by reminding me how nice it is when the collective psyche is focused on fun things 😂

This lil’ guy is living my dream of being hidden in a shrub, nibbling on flowers — not thinking about computer science/ world affairs 😂 On the flip side, I don’t live in mortal fear of the dogs at the dog park.

Visiting Chicago, god this place is gorgeous.

The art institute has a pretty broad exhibit on Symbolism. If it were me, I would’ve titled it “Brooding, but sometimes colorful.”

I am recruiting PhD students to start in 2026! If you are interested in robustness, training dynamics, interpretability for scientific understanding, or the science of LLM analysis you should apply. BU is building a huge LLM analysis/interp group and you’ll be joining at the ground floor.

I’ve seen this movie before! That fog grows, creeps closer, before revealing what it contains.

Work has gotten so busy that I’ve resorted to inventing fake Guru Laghima quotes to describe my life, Zaheer-style. “Let go your earthly tether to interpretability. Achieve 0.001% improvement. Become wind.” – Guru Laghima

I’m sitting at a cafe where it’s all tech workers with paper notepads next to their laptops. And I can’t help but feel like working from a cafe would attract the kind of tech worker who would enjoy writing things on nice stationary. 🤣

A ton of attention over the years goes to plots comparing open to closed models. The real trend that matters for AI impacts on society is the gap between closed frontier models and local consumer models. Local models passing major milestones will have major repercussions. buff.ly/ccMJydQ

Someone get ucla law on the line, cause this data viz’s color choice is hot garbo.

If it provides any comfort/ assuages imposter syndrome, I just learned about a very smart person who wrote pedabytes of NaNs, bc of a bug in their pipeline.

The UC Berkeley School of Information is hiring an assistant professor in the broad field of Information--including areas of info seeking/retrieval, digital humanities, cultural analytics, info viz, & philosophy of information (among others). Deadline Nov 1! aprecruit.berkeley.edu/JPF05014

They were selling these in the gift shop in the Seattle art museum. And, my god, the restraint I showed? Applaudable. *waits for applause* But also, these are absolutely stunning.

“You get 100 flops per minute!” “Floating point operations?” “What? Pfft. No. Fish- lopped per minute!”

You might think tpus/GPUs are complicated, but have you tried operating a 1910 “fish processing machine”?

Somewhere, my Korean and Chinese ancestors are like “oh yeah? You like ordering Japanese things? GET TARRIFFED, SON!”

Comparing AI Labs in the United States and China, by Jingyuan Liu

🔹Multi-Token Prediction → turbo-charged speculative decoding Demo: chat.qwen.ai Blog: qwen.ai/blog?id=4074... Model - Huggingface: huggingface.co/collections/... Model - Kaggle: www.kaggle.com/models/qwen-... Alibaba Cloud API: www.alibabacloud.com/help/en/mode...

I’ve lost power in Seattle for like the third time in a month. Which is not helping my nostalgia for Chicago’s better transit & stable power grid 😂 Otoh, there are some pretty church buildings in the area.

Super fun high-level talk on the history of ml frameworks. 18:50 is a nice overview of torch compilers 19:10: has a nice demarcation of user/execution categories 21:30 nicely describes, why GPUs? Nicely goes between broader historical effects, software, and hardware. youtu.be/139UPjoq7Kw?...

I really enjoy working from cafes, and zero trust network access/ ip reputation filtering/ geofencing is going to ruin me 😂

There are some perks of my job switch…but having just visited the Seattle Riot office — work somewhere with a dope IP, the work spaces are so great

At long last, I can post my team's summer project: applied modules to teach how ML/AI tools are changing social science and humanities research: ubcecon.github.io/praxis-ubc/ Highlights: LegalBERT to analyze anti/pro-immigrant sentiment in 19th c. BC law: ubcecon.github.io/praxis-ubc/d... 🧵1

+1 I just think that a field that has both coded and taught young readers/writers for years — thrives in an era when ml papers fumble about their “rephrasing” strategies and preference datasets have not been treated with some basic stylometry.

Seattle, advertising itself to be the setting of many fog-related thrillers.

I thought I ordered Chinese food…but I might be receiving way more than I bargained for….

Just made a playlist called “The Goats of Melancholy,” and I really wish I had the musical talent needed to build a band on that name.

Now look, I can understand, just because transparent does not mean weak. But my primate brain, otoh, will not step on this.

"On the Theoretical Limitations of Embedding-based Retrieval" Paper: arxiv.org/abs/2508.21038 Repo: github.com/google-deepm...

💯 the artists/ designers at riot games should really take pride that they were heavily influential in this

Lit period surveys courses are great prep for tech. You’re bombarded with important sounding nouns that seem to be important for a few years before it’s time for the next cohort. “So Shelley pipes to a Keats container, but during Hazlitt we need to rollback to the Lake Poets build…”

Folks in tech have some very elaborate knowledge storing methods. Meanwhile, like a llm, I am like “yeah well, what if I just try to memorize all the training data? Watch me scale my parameters on command.” *gets headache*

“First day back, into the breach once more and scale those dot products.” - Shakespeare, or something

Oh dear god

Now that school is starting for lots of folks, it's time for a new release of Speech and Language Processing! Jim and I added all sorts of material for the August 2025 release! With slides to match! Check it out here: web.stanford.edu/~jurafsky/sl...

🚨🚨JB Pritzker confirms knowledge of Nomadic Warriors for Pritzker, affirms "Greater Illinois 🚨🚨🚨

Singing along with strangers at a trio of animated-singing heroes? Pretty great experience — I hope it happens more often. It was very funny that the two loudest songs sung by the theater were Golden & Soda Pop.

TFree-HAT 7B A new tokenizer-free LLM. Built on Hierarchical Autoregressive Transformer (HAT) architecture, these models achieve top-tier German and English performance while processing text on a UTF-8 byte level.

github.com/AnnotatedLMs... Just organizing different llm architectures into a single place for my own reference. The implementations are very approximate, sandbox-y, and over-organized. Meant to be the code ver of an “annotated bibliography.”