Desi R Ivanova
banner
desirivanova.bsky.social
Desi R Ivanova
@desirivanova.bsky.social
Research fellow @OxfordStats @OxCSML, spent time at FAIR and MSR
Former quant 📈 (@GoldmanSachs), former former gymnast 🤸‍♀️
My opinions are my own
🇧🇬-🇬🇧 sh/ssh
Pinned
Decent 0-shot performance (conditional on it has been like 100 years since retirement) 😅
Reposted by Desi R Ivanova
🏢 Vacancy: Novo Nordisk Postdoctoral Research Fellow (4 posts)
📍 Department of Statistics, University of Oxford
📃 Contract: Full time, fixed-term for 3 years
💷 Salary Range: £36,024 – £44,263
⏲️ Deadline: 12pm UK, 30 Apr 2025

Full details & how to apply 👉 shorturl.at/3l47e
April 7, 2025 at 9:56 AM
Our paper on the best way to add error bars to LLM evals is on arXiv! TL;DR: Avoid the Central Limit Theorem -- there are better, simple Bayesian and frequentist methods you should be using instead.

We also provide a super lightweight library: github.com/sambowyer/baye… 🧵👇
March 6, 2025 at 3:00 PM
Reposted by Desi R Ivanova
Come work with me!
We are looking to bring on more top talent to our language modeling workstream at @ai2.bsky.social building the open ecosystem. We are hiring:
* Research scientists
* Senior research engineers
* Post docs (Young investigators)
* Pre docs

job-boards.greenhouse.io/thealleninst...
The Allen Institute for AI
job-boards.greenhouse.io
February 25, 2025 at 1:07 AM
I’m teaching a grad course for the first time (a bit terrifying 😅) and I’ve decided to write a short blog post after each lecture, that will highlight a key takeaway from it and reflect on what can be improved.

First one ⬇️
February 21, 2025 at 1:10 PM
Reposted by Desi R Ivanova
📣 Jobs alert

We’re hiring postdoc and research engineer to work on UQ for LLMs!! Details ⬇️

#ai #llm #uq
Postdoctoral fellowships and research engineer positions available for an Oxford+Singapore project on uncertainty quantification in LLMs!

docs.google.com/document/d/1...

Oxford deadline is Feb 26. Pls apply if interested, forward to your contacts, contact me if you have questions 🙏🙏
LLM UQ Singapore+Oxford job ad
Postdoctoral Fellow and Research Engineer Positions on Uncertainty Quantification in LLMs in Oxford and Singapore By scaling up data, compute and model size, large language models (LLMs) have gained ...
docs.google.com
February 12, 2025 at 4:26 PM
📣 Jobs alert

We’re hiring postdoc and research engineer to work on UQ for LLMs!! Details ⬇️

#ai #llm #uq
Postdoctoral fellowships and research engineer positions available for an Oxford+Singapore project on uncertainty quantification in LLMs!

docs.google.com/document/d/1...

Oxford deadline is Feb 26. Pls apply if interested, forward to your contacts, contact me if you have questions 🙏🙏
LLM UQ Singapore+Oxford job ad
Postdoctoral Fellow and Research Engineer Positions on Uncertainty Quantification in LLMs in Oxford and Singapore By scaling up data, compute and model size, large language models (LLMs) have gained ...
docs.google.com
February 12, 2025 at 4:26 PM
A tiny "embers of autoregression" artifact in simple arithmetic

probapproxincorrect.substack.com/p/embers-of-au…
February 8, 2025 at 9:05 PM
Reposted by Desi R Ivanova
🌟 Research Collaboration Morning 🌟

Our recent Research Collaboration Morning featured talks from Hrushi Loya (SGBE) & @desirivanova.bsky.social (CSML) over tea & coffee, followed by amazing treats from Waste2Taste & Blakes Food. Thanks to Prof Simon Myers and Bev Lane for organising!
February 6, 2025 at 9:48 AM
Reposted by Desi R Ivanova
📢 My team at Meta (including Yaron Lipman and Ricky Chen) is hiring a postdoctoral researcher to help us build the next generation of flow, transport, and diffusion models! Please apply here and message me:

www.metacareers.com/jobs/1459691...
Postdoctoral Researcher, Fundamental AI Research (PhD)
Meta's mission is to build the future of human connection and the technology that makes it possible.
www.metacareers.com
January 6, 2025 at 5:37 PM
Reposted by Desi R Ivanova
We're excited to announce the ICML 2025 call for workshops! The CFP and submission advice can be found at: icml.cc/Conferences/.... The deadline is Feb 10. Submit some creative proposals!
ICML 2025 Call for Workshops
icml.cc
January 3, 2025 at 2:19 AM
Reposted by Desi R Ivanova
It also makes for a bad equilibrium where publicly disagreeing with a paper (except very mildly and boringly) is infrequent and therefore an extreme signal.

So people don't do it even when they feel published papers are wrong or misleading.

3/
December 24, 2024 at 2:44 PM
Being able to do “academic research” at a tech salary was a zero interest rate phenomenon
I have been saying for years: getting into deep learning to chase high-salary AI jobs where you can do what you want is a fool’s errand.

There’s a lot of competition, differentiating yourself except by luck or connections is hard, and most companies want direct benefit from the salaries they pay.
feeling a but under the weather this week … thus an increased level of activity on social media and blog: kyunghyuncho.me/i-sensed-anx...
December 23, 2024 at 6:17 AM
Reposted by Desi R Ivanova
For my first post on Bluesky .. I'll start by announcing our 2025 edition of EEML which will be in Sarajevo :) ! I'm really excited about it and hope to see many of you there. Please follow the website (and Bluesky account) for more details which are coming soon ..
Hello Bluesky! 🦋

This will be the official account of the Eastern European Machine Learning (EEML) community.

Follow us for news regarding our summer schools, workshops, education/community initiatives, and more!
December 15, 2024 at 6:39 PM
Reposted by Desi R Ivanova
cool now do this

bsky.app/profile/labu...
here's a Sora generated video of gymnastics
December 11, 2024 at 6:35 PM
Reposted by Desi R Ivanova
Prior specification is one of the hardest tasks in Bayesian modeling.

In our new paper, we (Florence Bockting, @stefanradev.bsky.social and me) develop a method for expert prior elicitation using generative neural networks and simulation-based learning.

arxiv.org/abs/2411.15826
Expert-elicitation method for non-parametric joint priors using normalizing flows
We propose an expert-elicitation method for learning non-parametric joint prior distributions using normalizing flows. Normalizing flows are a class of generative models that enable exact, single-step...
arxiv.org
November 26, 2024 at 7:11 AM
Reposted by Desi R Ivanova
I've never, personally, had any direct utility from traditional academic peer review for my scientific work.

I've never used that peer review as any kind of signal or credential in judging a paper or deciding whether to read it.

Nearly all the papers I read are preprints.
November 25, 2024 at 2:01 AM
How about we stop writing and submitting papers that should have been a blog/reddit post at best
The review debate has entered 🦋. Let me re-propose my favorite idea so far: We reduce everything to a single point of entry with rolling review that only checks for scientific correctness (like TMLR). Once a paper is published there, you can apply for it to be presented at an upcoming conference.
We need to abolish peer-review conferences, at the very minimum. Everyone knows it’s a charade, yet they still keep playing the game.
November 24, 2024 at 10:45 AM
Reposted by Desi R Ivanova
I was raving about this paper for 1 month or so, I think everyone working on Reinforcement Learning should read it, and probably people working in other machine learning subfields should read it too
Adding my love letter to

arxiv.org/pdf/2304.01315

Empirical Design in Reinforcement Learning
by
Andrew Patterson, Samuel Neumann, Martha White, Adam White

JMLR 25 (2024) 1-63
#ReinforcementLearning

These aren’t the heroes we deserve, but they are the heroes we need.
arxiv.org
November 23, 2024 at 6:20 PM
Reposted by Desi R Ivanova
For those who don’t feel like they fit into my Grumpy Machine Learners list (which I still need to update based on 100+ requests) I’ve created another starter pack:

go.bsky.app/Js7ka12

(Self) nominations welcome.
November 22, 2024 at 6:40 PM
Reposted by Desi R Ivanova
The beta version of BayesFlow 2.0 is becoming more powerful and stable by the day. If you are curious about Amortized Bayesian Inference, give BayesFlow a try!
github.com/bayesflow-or...
GitHub - bayesflow-org/bayesflow at dev
A Python library for amortized Bayesian workflows using generative neural networks. - GitHub - bayesflow-org/bayesflow at dev
github.com
November 22, 2024 at 8:52 AM
Bring stats to LM evals!!

open.substack.com/pub/desiriva...
Inbox | Substack
open.substack.com
November 21, 2024 at 4:59 PM
💥 the most exciting new CDT just opened for admissions!!

More details will be announced soon, I promise it will be huge!!

www.dtc.ox.ac.uk/foai-cdt
Fundamentals of AI Centre for Doctoral Training
www.dtc.ox.ac.uk
November 21, 2024 at 12:47 PM