Jenny Qu @GuanniQu
safely merging w machines • previously 0days via RL @Pebble_bed • ctf player @SquidProxyLover • RL for math @caltech • she/they • DMs open quguanni.com sf Joined November 2020-
Tweets802
-
Followers2K
-
Following1K
-
Likes3K
i'd love git and a build system for agent beliefs. git to track which claims changed; a build system to flag every decision that depended on them for re-evaluation
wondering how much alignment is just disagreement below the user's correction threshold, and what happens when a consequential decision needs a distinction finer than that threshold, since acceptance is a noisy label for intention. sometimes it just means “not worth correcting”
a model update can break a security assumption without changing a single API. yesterday the model couldn't connect two harmless-looking tools into a dangerous capability. today it can.
the dangerous part of sharing memory between agents is that a low-privilege agent can write a belief that a high-privilege agent later executes
realizing a tool can be read-only and still increase an agent's effective privileges. reading the deployment guide can teach it how to use credentials it already had
and if BCI adapts to how i correct it, and i adapt to which mistakes are easiest to correct, we could converge on a language neither of us would have chosen alone. calibration as accidental language invention
wait so a BCI completing my thoughts can help form the intention i'm supposed to judge it against, and i can reject the suggestion but can't unhear it. make that loop fast enough and .. we are one and thinking them together. 🔗
used to treat thms like black boxes and pull in the proofs lazily to skip prereqs. now i do it to my whole workflow and the proofs never come. i'm just f_want(x) and "looks good i guess". Merge Me With The Model Please, give me the middle of 🥪
and ai can now arrange a thousand matches a day, the maintainers are the parents who have to bless each one individually
i’m realizing a lot of vulnerability research is matchmaking. i find two components that disagree about what the same bytes mean and introduce them to each other. unfortunately they have chemistry, fall in love, and eventually agree with each other
@drfeifei and x.com/guanniqu/statu…
1/ We shouldn’t measure a math problem’s value by how many steps it takes to solve. A lot of “hard” contest problems will vanish as search and hardware improve. The problems that matter are the ones whose solutions force a new way of thinking.
@drfeifei been thinking abt this for a while, and this is my attempt at answering: x.com/guanniqu/statu…
Wondering how we can make LLM understand the concept of mathematical beauty. If A = B then the further away A is from B the more beautiful a theorem. But some argue that the perception and experience of beauty is personal and not transferable.
CTFs are security’s best excuse to learn by breaking things together. We need to resurrect them !! 🪦🧩
what if ai helps the universe’s bubbles fuse before they pop and accidentally creates a new mind? im excited to meet it, also slightly worried im already in its group chat, and the three dots typing indicator is me
maybe a latent thought corresponds to an equivalence class of meaning-preserving representations, and communication selects from the subset the sender can express and the listener can decode. so 10 bits/s may measure the rate of projecting thought into behavior, and not the thought’s intrinsic bitrate.
oh woww, if this holds, functional-equation symmetry also clears \Re s<1/8. every Dirichlet L-function’s nontrivial zeros are trapped in [1/8,7/8].
Wait, the wild part is every Dirichlet L-function, not just \zeta. If correct, that rules out Siegel zeros past 7/8 and gives an x^{7/8+\epsilon} prime-counting error in each fixed progression. How are they controlling the conductor?
Quasi-RH?!?!???! Are you kidding me? If a human did this, it would be an instant Fields Medal, no questions asked. RH says zeta has no zeros in Re(s)>1/2. The best we had until a second ago was a region that got thinner and thinner the higher up the imaginary axis you go. I
wish usb-c were color coded like usb-a... please just let the plastic tongue leak one bit of type information
🫡 prof romik
Thoughts on math and AI. Are human mathematicians redundant in a world in which AIs are better than humans at proving theorems? The answer may surprise you. blog.danromik.com/the-feedback-l…
Jędrzej Maczan @ �... @jedmaczan
656 Followers 363 Following @CUBoulder, tiny-vllm, @PyTorch. Art by https://t.co/rs0nMCNnN6
Patricia @jnm38dPatricia
66 Followers 402 Following Curating useful AI tools for builders and creators.
Schrodinger's Cockato... @MaxDiffusionRL
905 Followers 4K Following Anti-Goodheart'er. Likes surplexity/epiplexity/higher order interactions/surprise. Wealth of weak ties. Tensegrity-lover
Suha @suhackerr
815 Followers 973 Following Disclaimer: The views expressed here are solely my own and do not reflect the views of my employer or any organization I am or was affiliated with
saanvi @saanvisub
16 Followers 521 Following cs + neurobio @caltech / prev @stanfordmed, @seattlechildren life is programmable! ml for proteins + neuro part-time musical poet and writer
0xBvlgari @0xbvlgari
9K Followers 10K Following
jia @jia_seed
21K Followers 7K Following 22. prev. drone research, dropped big tech currently internet data as co-founder @ jam
Sky Grip @skygrip777
15 Followers 2K Following
John Doe @sprinklepie8uh
17 Followers 259 Following This account has been hacked. I've reported it to authorities.
vishnuk @varmanpwp
20 Followers 222 Following trying to make coding agents less annoying to debug with(mood rn) npm i -g @varman96/paper
Jonathan Görtz @jonathan_gortz
1K Followers 1K Following Transhumanist. YC W26. @Starcraft grandmaster
MuRo @M_Sumaraq
161 Followers 7K Following
monkeybizniz @monkeymktplace
142 Followers 1K Following
Magnus Maximus @maximus_mag
9 Followers 323 Following Progressive Realist. Sometimes I strive to be as hawkish as the most hawkish liberal internationalists.
Noam Shalev @noamshh
14 Followers 144 Following
Vi @VIREseaon
1 Followers 16 Following
Alan O'Donnell @alanodonnell
15 Followers 1K Following
Mohammad Alaggan, Ph.... @malaggan
2K Followers 5K Following Principal Engineer at @AWSCloud (CloudFront). Opinions are my own.
Michael Kheifetz @michaelkheifetz
45 Followers 2K Following
Theis Rauhe @trixOcto
52 Followers 1K Following
Onur @onurtnuc
4 Followers 199 Following
Alex W @WAlex99501
16 Followers 433 Following
Lev @chumwonktonodI
256 Followers 3K Following Most of the time, I try to be less confused than yesterday and fail. PhD student in physics @FU_Berlin. Always happy to chat and learn more.
Brian Huang @brianryhuang
8K Followers 3K Following midtraining @GoogleDeepmind | prev math and cs @mit
randomfunandnonsense @randomfunch
91 Followers 1K Following
James @jaouwest
31 Followers 2K Following
Ahmed Youssef @QuantumRamses
246 Followers 3K Following Theoretical physicist. Uncertainty and causality in ML. VP Data Science. Berlin. Opinions are my own.
Rulphblack @VictorHerediaM2
18 Followers 194 Following Ciudadano del cosmos, empedernido melómano del conticinio, aspirante a ensoñador desadaptativo con toques de curiosidad epistémica.
Burny - Effective Cur... @burny_tech
20K Followers 11K Following On the quest to understand the fundamental mathematics of intelligence and of the universe with curiosity. https://t.co/mMchI2d4pg Upskilling by @StanfordOnline
Kyle Sargent @KyleSargentAI
2K Followers 1K Following I am a researcher @openai. I work on post-training, improving multimodal reasoning + perception capabilities | Prev: CS PhD @Stanford w/ @jiajunwu_cs @drfeifei
🦖 @KarlMarxSweetie
24 Followers 1K Following
Noe JeFF @TheRealNoeJeFF
24 Followers 365 Following
Nguyen Thai Phong @phnguyenthai
12 Followers 788 Following
rhombitech @rhombitech
3 Followers 82 Following
MathStunned @MathStunned
8 Followers 1K Following
Michael Rose @mrose00
176 Followers 6K Following
Farzad 🇺🇸 🇮�... @farzyness
400K Followers 2K Following Analyzing Disruptive Innovation Abundance or Collapse: https://t.co/IZNjdFyEGK Master Plan: https://t.co/Ujh87zWw4b
jia @jia_seed
21K Followers 7K Following 22. prev. drone research, dropped big tech currently internet data as co-founder @ jam
CodeGraff @trycodegraff
52 Followers 6 Following An AI that actually does the work. Code, research, and automation. Give it a task. Let it cook.
Patrick Mineault @patrickmineault
24K Followers 3K Following NeuroAI researcher @ Amaranth Foundation, safety, open science. Previously engineer @ Google, Meta, Mila.
Jasmine Wang @j_asminewang
15K Followers 1K Following formerly safety @OpenAI @AISecurityInst opinions very much personal!
Kim Jinu @uz56764
1K Followers 160 Following https://t.co/gf216pulU0 / KAIST 24 / DEFCON CTF 2023~2024 Finalist / The Amazing Digital Orange
TrendAI Zero Day Init... @thezdi
90K Followers 18 Following TrendAI Zero Day Initiative™ (ZDI) is a program designed to reward security researchers for responsibly disclosing vulnerabilities.
Michael Lin, MD PhD �... @michaelzlin
29K Followers 457 Following Prof. of Neurobiology & BioE @Stanford @MichaelLinLab ☘️🧪🦠🧠🔬📖🇺🇲🌏 Neurons, viruses, medicines. Invented predecessor of nirmatrelvir. Bad manners blocked.
Acer @AcerFur
11K Followers 2K Following Furry pure maths student @Cambridge_Uni. | 🇬🇧🇵🇹 22 He/Him | prev: @OpenAI | Opinions are my own.
Dave Munichiello 🇺... @davemuni
4K Followers 1K Following Managing Partner @GVteam where we empower high-agency, high-IQ founders. https://t.co/igCXzcaDbZ. Prev Prod/GtM/Ops @KivaSystems+@Amazon. Paratrooper+Combat Veteran.
Nathan Lambert @natolambert
106K Followers 954 Following Open model research, founder, director @trillium_labs. Prev. co-led Olmo at Ai2. Writes @interconnectsai, wrote https://t.co/alRXKINTwE
AIR @air__security
503 Followers 6 Following The Context Firewall for AI Agents AIR sits between agents and the outside world, continuously analyzes and filters every input into an agent’s context.
the tiny corp @__tinygrad__
84K Followers 209 Following We make tinygrad; sell tinybox for the GPU middle class. Our mission is to commoditize the petaflop.
Sarah Zeng @sarahzengy
663 Followers 348 Following chief of staff @braintrust. angel investor @snowyleopardvc. prev applied math @harvard
Eric Glyman @eglyman
395K Followers 2K Following Co-Founder at Ramp (@tryramp). New York City. Previously co-founded Paribus (Acq. by Capital One).
Umesh Khanna 🇨🇦... @forwarddeploy
15K Followers 5K Following talent eng @spacexai | 👼 forward deployed angels | 2x founder | @uwaterloo
Jeff Dean @JeffDean
531K Followers 6K Following Co-founder & CEO of Discovery Loop. Former Chief Scientist, Google. Helped build many Google products, TPUs, Gemini, TensorFlow, MapReduce, Bigtable, ...
Yurii Rosso @YuriiRosso
1 Followers 15 Following Revolutionizing the tourism industry from Singapore to the world. Granted by Singapore Tourism Board. 2024 ILTM Top speaker. 2024 TTG Asia Top travel adviser.
Azalia Mirhoseini @Azaliamirh
22K Followers 646 Following Founder @RicursiveAI, Asst. Prof. of CS at Stanford. Prev: DeepMind, Anthropic, Brain. Co-Creator of MoEs, AlphaChip, Test Time Scaling.
Anna Goldie @annadgoldie
12K Followers 162 Following Founder & CEO at @RicursiveAI. Prev: @GoogleDeepMind, @AnthropicAI, @Stanford PhD, @MIT. AlphaChip co-lead.
Ekin Dogus Cubuk @ekindogus
10K Followers 499 Following Co-Founder of @periodiclabs Past: Lead of materials science and chemistry at @GoogleDeepMind; Google Brain
Liam Fedus @LiamFedus
42K Followers 1K Following Building industrial-scale science at @periodiclabs Past: VP of Post-Training @OpenAI; Google Brain
rohan anil @_arohan_
46K Followers 2K Following co-founder, board @coreauto - and continuing to aspire to understand deep learning.
Jerry Tworek @MillionInt
43K Followers 1K Following CEO and co-founder of @coreauto former VP of RL @ OpenAI : reasoning models, o3, o1, GPT4, ChatGPT, Codex, RL for robots cautious AI optimist
Rach @rachpradhan
3K Followers 241 Following @nusingapore '24 | databases → RL infra → financial world models | 12++ solo hack wins (CalHacks,HTN etc.) | angel investor | i like making fast things
Phil Hedayatnia @PhilHedayatnia
7K Followers 2K Following cofounder & designer @airfoilstudio ✨ prev @spc projects to de-slop the web! https://t.co/vdfVgvqydQ ・ https://t.co/wgvye0zMrS ・ https://t.co/Y6YBM9POnt
rochelle shen @neuro_morphic
1K Followers 903 Following Biohacker girl waiting for her next upgrade. { previously ML @PalantirTech, comp neuro @Penn }
🧠kn𖡹t @neuralkn0t
823 Followers 1K Following
I. Yosun Chang @Yosun
4K Followers 1K Following {wonder, innovation, elegance} ∈ I turn emerging technologies into award winning apps. Ex-Hackathon pro. #3D #AR #AI since forever. Mad science and artistry ❤️
Sonia Joseph @soniajoseph_
17K Followers 1K Following co-founder of World Mechanics @world_mech_ai | prev world models / JEPA, interpretability @AIatMeta @Mila_Quebec @MATSprogram @Princeton
Grace Li @grx_xce
10K Followers 657 Following cofounder @intelligence_ai (@designarena, @predictionbench). Prev cs + neuro harvard, swe apple
terjanq @terjanq
11K Followers 271 Following security enthusiast that loves hunting for bugs in the wild. co-founder and player of @justCatTheFish. infosec at @google. opinions are mine.Trends for United States
You might like
