nick | @[email protected]

Very happy and excited that our paper "Isolating Functions at the Hardware Limit with Virtines" has been accepted to #eurosys22! Great work everyone!

Alright that's it. This stupid-ass site is just meta thirst posts to get the attention of a moronic billionaire. I'm deleting this app. For the one person who follows me, I'll be elsewhere 🫡

The most *nation* meal ever according to midjourney AI 1. Mexican

Just got off a call with Reddit about the API and new pricing. Bad news unless I come up with 20 million dollars (not joking). Appreciate boosts. reddit.com/r/apolloapp/co…

never forget what they took from you

@toxicxxfuzion Neither do I, but at least I can swipe around mindlessly. Mastodon is all things I have to think about lol

@awesomekling The video is running faster than the game does :)

Quick clip of Steam in the Ladybird browser 🐞🌍 ✅ From-scratch HTML/CSS engine ✅ From-scratch JavaScript engine ✅ From-scratch video codec Don't listen to defeatists. It is absolutely possible to build new browsers. 🤓🌄

@shwestrick @kerckhove_ts Though it takes more work to update the bloom filter when you remove items. There was work to have two bloom filters, one that tracks insertions, and one that tracks removals. Or you rebuild the bloom filter when you remove, or be fine with slowly increasing false positives

@kerckhove_ts take any set data structure and add a bloom filter to it. `member` just checks the underlying set directly. `notMember` checks the bloom filter before falling back on checking the underlying set. This will potentially be much faster.

What's the point of a function like this: notMember k m = not $ member k m Are there _any_ circumstances in which `member` and `notMember` are a partition but `notMember` is more performant, in order of magnitude, than `not . member`?

well. @UMich has officially docked all pay but $100 for april from its grad workers, putting many of us in a position of dangerous financial precarity. friends/allies, please consider donating to @geo3550 strike fund at givebutter.com/GEOStrikeFund. we need all the help we can get ✊

It's insane to me that people *pay* to follow a billionaire (Here, he is *paying* to follow his billionaire brother ...) Delusional.

There's a lot of folks under the misunderstanding that it's now possible to run a 30B param LLM in <6GB, based on this GitHub discussion. This is not the case. Understanding why gives us a chance to learn a lot of interesting stuff! 🧵 github.com/ggerganov/llam…

Won't lie, I'm sick of GPT

@toxicxxfuzion How about TLOU?

Had the landlord come by to check that my pet wasn't destroying the house. Within 5 seconds of walking into the house, they say "that dude couldn't hurt a fly" That dude:

Last October my department unexpectedly voted negatively against my tenure case. Since then, all of my students have had first author papers accepted - TACAS, OOPSLA, and now PLDI. I cannot begin to express how hard these last few months have been and how I feel today.

Another example of why artifact eval in computer science should be at the same time as review. It'a crazy that code is only ever made public *after* acceptance. Most computer science *is* the code, not just plots on a page, and the code should be critiqued as part of review

When a HPCA'23 paper claimed to break MIRAGE, we were excited to read! But their code has bugs - (1) Buggy cipher not randomizing well, (2) Starting from broken cache state. After bug-fixes, no set-conflicts! Breaking of MIRAGE is a mirage. Our analysis: tinyurl.com/bdzmfaav

Google Calendar, but it shows the cost of the meeting

after working on this terrible ML homework, I'm convinced nobody knows what they are doing in machine learning. Numpy is nonsensical.

I don't understand how anyone puts up with python's bs