Viktor Senderov retweeted
Well done to CRT students @SophiieMatthews and @kev_ryan_95 on winning prizes for best lightening talk and poster presentation respectively @UniGalwayMaths research day👏👏👏
You can see the ships lights all go out and then eventually some of them come back online shortly before the impact. I wonder if there was some sort of electrical issue on the ship, preventing it from maneuvering?
Viktor Senderov retweeted
it's actually crazy how much better triangle notation is than the standard notation
Viktor Senderov retweeted
The #OpenAI soap-opera hijacked the attention of millions of productive people and nonsensually crammed the fine details of the debate between #EffectiveAltruism and #EffectiveAccelerationism into them, a genuinely absurd debate that was allegedly at the center of the drama.
1/

ALT A giant glowing green Cartesian plane, receding into the distance. In the foreground are Tenniel's illustrations of Tweedledee and Tweedledum, pulling their hair in rage. Each has a glowing HAL9000 eye from Kubrick's '2001: A Space Odyssey' on his belly. In the background, Tenniel's Beamish Boy confronts the Jabberwock, whose eyes have also been replaced with HAL9000's eye. Image: Cryteria (modified) https://commons.wikimedia.org/wiki/File:HAL9000.svg CC BY 3.0 https://creativecommons.org/licenses/by/3.0/deed.en
Viktor Senderov retweeted
Do I have this right?:
“GPT store”
@adamdangelo pissed
Starts coup
@ilyasut convinced
Board listens
@sama fired
@gdb ousted
@miramurati CEO
Greg quits
Employees pissed
“We’ll quit”
Board scrambles
Bye Mira
Hello @eshear
@satyanadella negotiates
Hires Sam
Hires Greg
Ilya regret
Employees threaten
Letter drafted
747 sign
Shared trauma
Bond stronger
Netflix series
Emmy award
AGI achieved
The best way to become a 10x developer is to teach 10 developers to be as good as you. --- not me, but I like the quote
Viktor Senderov retweeted
ProbAI goes beyond Nordics! Apply now to the Tropical Probabilistic AI School 2024 in Rio de Janeiro 🇧🇷, 29 Jan – 2 Feb. tropical.probabilistic.ai 🌴 #ProbAI #ProbabilisticAI
Viktor Senderov retweeted
I still believe deeply in the need for a radical restructuring of science publishing to create an open and fair system that works well for science, for all scientists everywhere, and the public that supports us, and am proud of the steps we have taken to begin to get us there.
Viktor Senderov retweeted
TreePPL: A Universal Probabilistic Programming Language for Phylogenetics dlvr.it/SxQQf1
Viktor Senderov retweeted
We're all locally famous at #tdwg2023
Attending the LLM session at #ISMBECCB2023 . A lot of people share that they are ChatGPT to write outlines, but then fill in the details themselves. I think this is backwards: write your own bullet points and let LLM do the mundane work of forming syntactically correct sentences
Viktor Senderov retweeted
📢 BREAKING: the Biodiversity Knowledge Hub is here!
💜🧡💚
#BiCIKL_H2020 is proud to announce the arrival of this game-changer in scientific research.
Read the full story and TEST IT here:
🤜bicikl-project.eu/news/BKH🤛
#biodiversity #FAIRdata
Viktor Senderov retweeted
#PhD student: Makes $30k a year. Works on weekends as well. Zero life-work balance.
"Do you think I have a chance to become a professor?"
Prof: "Yes, of course! Finish this project and we will publish excellent papers. I am sure you will easily find a faculty position."
▫️
2 years later:
Student finishes the project. Professor writes a report. Papers are published.
Student: "Do you think my CV is strong enough?"
Prof: "Yes, you are the best!"
▫️
Next 4 months:
Student submits 50 well-tailored applications for faculty positions.
Zero interviews. A lot of broken dreams.
▫️
Key takeaways:
1. Make sure you distinguish encouragement from reality.
- By encouraging you, your advisor may unintentionally give you too much hope. Keep a cool head.
2. Always ask other faculties for external opinion on your case.
- Your advisor’s opinion is always biased. Look for more input outside your group.
3. Don’t expect fairness during candidate selection.
- Hiring process is subjective by definition. It is done by people with very different views on who is the best. You may put tons of efforts into a research statement only to find out later that no one really reads it.
4. The reality is brutal.
- Departments can receive 300-500 candidates per opening. Many have excellent CVs and cool ideas. At top- and mid-rank universities, selection criteria can become extremely questionable (like, who exactly was your PhD advisor? Is your recomm. letter 3 pages long? etc).
And there is no need to say “You don’t know anything about it. It’s not like this”.
I went through this myself. Many times. Along with many colleagues.
Do not expect fairness. See luck as a big factor.
Apply broadly but have a backdoor ready.
#AcademicTwitter
Viktor Senderov retweeted
A 14-line Python script using gzip outperforming a 345m parameter transformer model is probably the most hilarious result I've seen all year.
aclanthology.org/2023.findin…
Viktor Senderov retweeted
Thrilled to see our study led by @AlexDSkeels out now in @ScienceMagazine!
We show that asymmetries in the biotic interchange across Wallace’s Line were driven by differences in precipitation between Asia, Australia and the stepping stones in between.
science.org/doi/10.1126/scie…
#ISMBECCB2023 Does any know how to upload the PDF and MP4 of the poster ahead of the "posterpalooza"? The Juno system is slow and very hard to navigate :(
Viktor Senderov retweeted
Can LLMs generate mathematical proofs that can be rigorously checked?
We release LeanDojo (leandojo.org): an open-source playground consisting of toolkits, benchmarks, and models for LLMs to prove formal theorems in the Lean proof assistant.
Key features:
- Tools for data extraction and interacting with Lean.
- Fine-grained annotations of premises (e.g., existing lemmas) in proofs: where they are used and defined.
- LeanDojo Benchmark: 97K human-written theorems/proofs for developing machine learning models on theorem proving.
- ReProver (Retrieval-Augmented Prover): the first LLM-based prover augmented with retrieval for premise selection.
We open-source everything, providing the first set of open-source LLM-based theorem provers without any proprietary data, model, or code.




















