user avatar
Naomi Saphra
@nsaphra
Waiting on a robot body. All opinions are universal and held by both employers and family. Now a dedicated grok hate account. Accepting ML/NLP PhD students.
Boston
nsaphra.github.io
Joined November 2010
1,243
Following
10.6K
Followers
Posts
  • Pinned
    user avatar
    Naomi Saphra
    @nsaphra
    Mar 27, 2025
    Life update: I'm starting as faculty at Boston University in 2026! BU has SCHEMES for LM interpretability & analysis, so I couldn't be more pumped to join a burgeoning supergroup w/ @najoungkim @amuuueller. Looking for my first students, so apply and reach out!
    108K
  • user avatar
    Naomi Saphra
    @nsaphra
    Jul 20, 2025
    As confirmed by the new IMO rankings, Grok 4’s eye-popping benchmarks were driving by the following innovations: - train on test - train on test - train on test
    432K
  • user avatar
    Naomi Saphra
    @nsaphra
    Jun 15, 2023
    Regular reminder of the best mathematical resource in machine learning, The Matrix Cookbook. Don't know how anyone ever does any math without it. math.uwaterloo.ca/~hwolkowi/matr…
    218K
  • user avatar
    Naomi Saphra
    @nsaphra
    Jan 12, 2016
    My hobby: watching underpaid, overworked engineers sacrificing their 20s to an early stage startup ridicule people who buy lottery tickets.
  • user avatar
    Naomi Saphra
    @nsaphra
    Apr 14, 2016
    What idiot called it "deep learning hype" and not "backpropaganda"
  • user avatar
    Naomi Saphra
    @nsaphra
    Dec 9, 2020
    Why isn't color-coding explanations more common?
  • user avatar
    Naomi Saphra
    @nsaphra
    Jan 11, 2021
  • user avatar
    Naomi Saphra
    @nsaphra
    Nov 23, 2022
    Have you ever noticed how Chinese and American researchers both publish at #NeurIPS, go to the same conference---and then barely cite or talk to each other? @BingchenZhao @gu_yuling @in4dmatics & I have, and we'll be presenting on it at @AiCultures! arxiv.org/pdf/2211.12424…
    One Venue, Two Conferences: The Separation of Chinese and American Citation Networks Abstract At NeurIPS, American and Chinese institutions cite papers from each other’s regions substantially less than they cite endogamously. We build a citation graph to quantify this divide, compare it to European connectivity, and discuss the causes and consequences of the separation.
  • user avatar
    Naomi Saphra
    @nsaphra
    Oct 16, 2025
    I’m recruiting PhD students for 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.
    user avatar
    Naomi Saphra
    @nsaphra
    Mar 27, 2025
    Life update: I'm starting as faculty at Boston University in 2026! BU has SCHEMES for LM interpretability & analysis, so I couldn't be more pumped to join a burgeoning supergroup w/ @najoungkim @amuuueller. Looking for my first students, so apply and reach out!
    113K
  • user avatar
    Naomi Saphra
    @nsaphra
    Mar 4, 2021
    Did you know the Fisher Information Matrix is the second-order Taylor approximation ... to KL divergence??????????????????????????????????????????????? I'm shaking idk how to handle this. what a good fact
  • user avatar
    Naomi Saphra
    @nsaphra
    Jul 20, 2025
    Replying to @nsaphra
    tfw you could have joined a bleeding edge LLM lab but you were too desperate to train a substandard nazi waifu on the test set
    28K
  • user avatar
    Naomi Saphra
    @nsaphra
    Sep 4, 2023
    Just got a desk reject, post-rebuttals, for a paper being submitted to arxiv <30 min late for the anonymity deadline. I talk about how the ACL embargo policy hurts junior researchers and makes ACL venues less desirable for NLP work. I don’t talk about the pointless NOISE it adds.
    660K
  • user avatar
    Naomi Saphra
    @nsaphra
    Jan 26, 2021
    Finally at the stage of writing a PhD thesis where I get to settle 20-year-old grudges.
  • user avatar
    Naomi Saphra
    @nsaphra
    Nov 10, 2023
    It's not the first time! A dream team of @enfleisig (human eval expert), Adam Lopez (remembers the Stat MT era), @kchonyc (helped end it), and me (pun in title) are here to teach you the history of scale crises and what lessons we can take from them. 🧵arxiv.org/abs/2311.05020
    user avatar
    Andriy Mulyar
    @andriy_mulyar
    Mar 15, 2023
    Replying to @andriy_mulyar
    my Twitter feed is full of ph.d. students having an existential crisis
    125K