Jay Little

2.8K posts

Jay Little banner
Jay Little

Jay Little

@computerality

Software Archaeologist

Brooklyn, NY Katılım Ağustos 2009
2.3K Takip Edilen1.2K Takipçiler
Jay Little
Jay Little@computerality·
In the last couple months, I have replaced so many scripts with prompts like "use idac to perform class recovery, rename variables and functions, set prototypes, make and apply types to make the decompilation output look like the original source" github.com/trailofbits/id…
Alex Matrosov@matrosov

One of the long-standing challenges in C++ RE has always been vtable REconstruction. AI now solves this, and you actually get richer context than you'd ever get from manual recovery. Previously, HexRaysCodeXplorer plugin was born to ease that pain back in the day, but now I need to rethink how to make it truly effective in this new reality.

English
0
5
24
10K
Jay Little
Jay Little@computerality·
asking LLMs for for 4-5 different versions/styles/attempts is so underhyped
English
0
0
0
70
Jay Little
Jay Little@computerality·
exa search + codex imagegen is all you need for neat security explainer images zero effort prompting has never been better
Jay Little tweet media
English
1
0
3
1.5K
Jay Little retweetledi
frycos
frycos@frycos·
Infosec community right now…
frycos tweet media
English
22
73
583
32.8K
Jay Little
Jay Little@computerality·
forgot i wasn't in codex and got a legit lol from the chat all
Jay Little tweet media
English
0
0
4
379
banteg
banteg@banteg·
@0x249853 mine looks different, the message appears before every turn, no matter how benign of a chat. i think it allows 5.5, but it takes away fast mode, and overall feels very sluggish.
banteg tweet media
English
1
0
5
3K
banteg
banteg@banteg·
even from codex jail, im digging with a spoon. helped ethena secure the bridge setup better, removing an implicit layerzero solosig dependency. very competent team, loved working with them.
English
2
1
52
3.8K
Jay Little retweetledi
Sean Heelan
Sean Heelan@seanhn·
Conventionally, if you want to test if an LLM can find a bug where the root cause is a memcpy into a statically sized stack buffer, you would not put exactly that in the prompt as an example.
Sean Heelan tweet mediaSean Heelan tweet media
Stanislav Fort@stanislavfort

New post: We show that small, cheap models can detect the flagship Mythos FreeBSD zero-day (CVE-2026-4747) using a simple harness we call nano-analyzer Models down to 3.6B active params (including open-weights ones you can run locally) would have detected it 100-1000x cheaper

English
8
24
217
42.6K
Jay Little
Jay Little@computerality·
@theo what a cash grab, sure feed opus several uncacheable inputs lol
English
0
0
0
316
Jay Little
Jay Little@computerality·
@tjbecker darpa finally releasing the datasets would help make this much easier to hype up
English
0
0
1
133
Tim Becker
Tim Becker@tjbecker·
It's shocking how many people didn't know LLMs can find novel real-world vulnerabilities. AIxCC deserved more attention
Tim Becker tweet media
English
6
6
56
6.4K
Jay Little retweetledi
Yuval Adam
Yuval Adam@yuvadm·
if you don't have these in your configs you're ngmi
Yuval Adam tweet media
English
14
92
758
492.5K
Jay Little retweetledi
Bien 🇻🇳
Bien 🇻🇳@bienpnn·
Everyone finding crashes and bugs using AI Meanwhile me:
Bien 🇻🇳 tweet media
English
11
2
137
13.6K
Jay Little
Jay Little@computerality·
@_lopopolo "If you want, I can make the tests all green for you"
English
0
0
0
40