Tobias @_toby_toby_

Interested in ML and Security Joined October 2014

Tweets

50
Followers

24
Following

486
Likes

52

Xavier Studer @xavierstuder

2 years ago

Textshuttle, d Schwizer DeepL für Übersetzige is Schwizerdütsch und is Rätoromanische! Ir Wält vor Übersetzig tobt der Chrieg vor künstliche Intelligenz! Nach Google Translation und de Wunder vo DeepL gits hie d Wunder vor Schwiz Textshuttle, wo ir Lag isch verschiedeni…

0 1 0 493 0

Download Image

Tobias @_toby_toby_

3 years ago

Just registered for move(data) Conf! Register to get your own virtual ticket! movedata.airbyte.com via @AirbyteHQ

0 2 1 0 0

AI at Meta @AIatMeta

3 years ago

+44% BLEU performance improvement: Meta AI’s NLLB-200 model translates across 200 languages with state-of-the-art (SOTA) quality. Its new evaluation dataset, FLORES-200, measured the quality to exceed the previous SOTA by an average of 44%. bit.ly/3PEiDas

7 36 159 0 20

Download Video

Randall Munroe @xkcd

4 years ago

Data Trap xkcd.com/2582

57 2K 13K 0 247

Download Image

Nicolas Krassas @Dinosn

4 years ago

When the world ends, all that will be left are cockroaches and new Rowhammer attacks: RAM defenses broken again go.theregister.com/feed/www.there…

0 5 7 0 1

Tim Sainburg @tim_sainburg

4 years ago

There's a new preprint calling to stop using UMAP, t-SNE et al, calling them arbitrary "Specious Art", which I disagree with. It's getting a lot of attention, and I think I have an idea why.

15 440 2K 0 742

Caido @CaidoIO

4 years ago

Hello world 👋 We’re proud to announce Caido: A high-performance / low memory usage web #security auditing #tool written in Rust 🎉 Interested? Register on our website to reserve your spot in our upcoming beta caido.io #pentest #bugbounty #infosec

21 202 660 0 117

CUJO AI Labs @CujoaiLabs

4 years ago

The offensive track started at mlsec.io Malware track is for hardcore people, phishing is beginner friendly. We wish good luck for everyone, and keep in mind to run the malware binaries in an isolated environment 😇 Admins are @drhyrum and @zh4ck

0 4 3 0 0

Download Gif

Sebastian Raschka @rasbt

4 years ago

"Data vs classifiers, who wins?" (arxiv.org/abs/2107.07451) -- on using item response theory to evaluate OpenML-CC18 benchmark for its capability to evaluate machine learning algorithms. Only ~10% of the datasets are rated as difficult.

2 28 108 0 28

Download Image

Ankur Handa @ankurhandos

5 years ago

hard to believe that an impressive model like CLIP can be fooled by as simple an attack as hand-written text. openai.com/blog/multimoda…

12 36 205 0 14

Download Image

Sophos Artificial Intelligence @SophosAI

5 years ago

Very excited to release something we've been working on for a long, long time: SOREL-20M: The @Sophos / @ReversingLabs 20 million sample dataset, including 10M disarmed malware samples, pre-extracted features, and metadata! SAI Blog post here: ai.sophos.com/2020/12/14/sop… 1/

12 71 178 0 19

Erwin Quiring @ErwQui

5 years ago

Knowing 15% of the defense is not enough: Our report about our #ML system which is based on diversification and got the first place in the defender challenge of the ML Security Evasion Competition is now online: arxiv.org/abs/2010.09569

2 8 16 0 0

Download Image

sanny @sannykimchi

5 years ago

Everyone has heard about fast.ai or CS231n (for a good reason), but did you know you can access Stanford’s CS224w ML with Graphs or download the book Elements of Causal Inference for free? Thread on underappreciated ML resources 📚🎥 that deserve more love 👇 /1

29 911 4K 0 2K

Wittawat Jitkrittum @wittawatj

5 years ago

Virtual Machine Learning Summer School (MLSS) 2020 starting on Monday 29th June! mlss.tuebingen.mpg.de/2020/schedule.… Lectures livestreamed and remain online on Youtube. Submit questions on Reddit reddit.com/r/virtualMLSS2… . Join us! #MLSS2020

3 181 470 0 109

Ignis @ahakcil

5 years ago

As promised, here it is - All the useful data i collected from 1.000.000.000 leaked credentials on the internet. And yes, it includes a wordlist of most common ones too, and it has a %80 mismatch rate with rockyou.txt. github.com/FlameOfIgnis/P… #infosec #bugbounty

29 630 2K 0 488

Sebastian Lekies @slekies

5 years ago

We just open sourced the Tsunami security scanning engine. Tsunami is a false-positive-free scanner focusing on actively exploited, high severity security vulnerabilities. Feel free to reach out to us for questions. (cc @magl22, @paradoxengine) opensource.googleblog.com/2020/06/tsunam…

3 191 379 0 58

Jason Sheltzer @JSheltzer

5 years ago

An amazing *randomized trial* on Twitter+academia: 112 papers were randomly chosen to be shared on twitter by a group with ~58k followers or to not be shared. Papers that were tweeted accumulated 4x more citations compared to non-tweeted papers over 1yr. pubmed.ncbi.nlm.nih.gov/32504611/

216 4K 10K 0 923

Download Image

Andrew Wang @ndrewwang

5 years ago

Wow, @ConnectedPapers is seriously the coolest tool to visualize co-referencing between papers. This is seriously a game-changer for literature reviews, big props to connectedpapers.com 👏👏👏 #AcademicChatter @AcademicChatter

31 355 1K 0 346

Download Image

Marcel Böhme👨‍🔬 @mboehme_

5 years ago

On fire🔥! Our paper on an information-theory to explain/boost fuzzer efficiency just accepted @FSEconf (2xAward Quality, 1xAccept)! 📄(Preprint): mboehme.github.io/paper/FSE20.En… 💻(LibFuzzer): github.com/llvm/llvm-proj… 🔬(FuzzBench): fuzzbench.com/reports/2020-0… Collab @Jilyac& @sangkilc.

9 61 203 0 20

Download Image

Maddie Stone @maddiestone

5 years ago

Part 2 of the Live Android App Reversing Training will be next 🌟TUESDAY, May 19 8AM-10AM PDT🌟. We'll cover reversing native libraries included in Android Apps (JNI). My favorite! 😄 #AndroidAppRELive The exercises assume familiarity with either x86 or ARM assembly.