Foivos I. Diakogiannis
@foivosdiak.bsky.social
X-tronomer, Senior Research Scientist (data scientist) - @ Data61/CSIRO. I mostly tweet about AI. Opinions are my own.
Pinned
GitHub - feevos/tfcl: Official repository for PTAViT3D and PTAViT3DCA models for field boundaries detection using S2 and/or S1 imagery.
Official repository for PTAViT3D and PTAViT3DCA models for field boundaries detection using S2 and/or S1 imagery. - feevos/tfcl
github.com
1/n 🚀 Just dropped our latest work on 'Temporal Fusion for Cloudy Imagery' - tackling challenges of clouded remote sensing data with cutting-edge AI.
🛰️🖥️ Check it out here: github.com/feevos/tfcl
🌍 #RemoteSensing #AI #DeepLearning #Geospatial
🛰️🖥️ Check it out here: github.com/feevos/tfcl
🌍 #RemoteSensing #AI #DeepLearning #Geospatial
Reposted by Foivos I. Diakogiannis
This is some great work!! I personally feel that one of the bottlenecks with memorization evals is having access to the gigantic training data. Super cool to see we can still run reliable evals without having access to the training data!
Want to know what training data has been memorized by models like GPT-4?
We propose information-guided probes, a method to uncover memorization evidence in *completely black-box* models,
without requiring access to
🙅♀️ Model weights
🙅♀️ Training data
🙅♀️ Token probabilities 🧵 (1/5)
We propose information-guided probes, a method to uncover memorization evidence in *completely black-box* models,
without requiring access to
🙅♀️ Model weights
🙅♀️ Training data
🙅♀️ Token probabilities 🧵 (1/5)
Information-Guided Identification of Training Data Imprint in (Proprietary) Large Language Models
High-quality training data has proven crucial for developing performant large language models (LLMs). However, commercial LLM providers disclose few, if any, details about the data used for training. ...
arxiv.org
March 23, 2025 at 11:24 PM
This is some great work!! I personally feel that one of the bottlenecks with memorization evals is having access to the gigantic training data. Super cool to see we can still run reliable evals without having access to the training data!
Reposted by Foivos I. Diakogiannis
On the subject of people traveling to the US right about now: Fuck, if I were from another country and I didn't need to come, I sure as hell wouldn't, the current administration has made it clear traveling here isn't safe, and why not take them at their word, there's a whole world to visit
March 16, 2025 at 5:53 PM
On the subject of people traveling to the US right about now: Fuck, if I were from another country and I didn't need to come, I sure as hell wouldn't, the current administration has made it clear traveling here isn't safe, and why not take them at their word, there's a whole world to visit
Reposted by Foivos I. Diakogiannis
STAND UP FOR SCIENCE SPEAKER ANNOUNCEMENT! ☀️
Bill Nye will be speaking at Stand Up for Science in D.C. on March 7th!
More speaker announcements coming soon, stay tuned! 👀☀️
Bill Nye will be speaking at Stand Up for Science in D.C. on March 7th!
More speaker announcements coming soon, stay tuned! 👀☀️
March 6, 2025 at 12:04 AM
STAND UP FOR SCIENCE SPEAKER ANNOUNCEMENT! ☀️
Bill Nye will be speaking at Stand Up for Science in D.C. on March 7th!
More speaker announcements coming soon, stay tuned! 👀☀️
Bill Nye will be speaking at Stand Up for Science in D.C. on March 7th!
More speaker announcements coming soon, stay tuned! 👀☀️
Reposted by Foivos I. Diakogiannis
PhD AI agents for more than most PhD humans cost.
Look forward to the sales figure on this, cc @edzitron.com
Look forward to the sales figure on this, cc @edzitron.com
March 5, 2025 at 4:34 PM
PhD AI agents for more than most PhD humans cost.
Look forward to the sales figure on this, cc @edzitron.com
Look forward to the sales figure on this, cc @edzitron.com
Reposted by Foivos I. Diakogiannis
February 28, 2025 at 11:14 PM
Reposted by Foivos I. Diakogiannis
Francis Collins, the NIH Director for 12 years, led the Human Genome Project and other NIH efforts for 32 years, resigned today. Key words from his resignation letter
www.nytimes.com/2025/03/01/u...
www.nytimes.com/2025/03/01/u...
March 1, 2025 at 6:07 PM
Francis Collins, the NIH Director for 12 years, led the Human Genome Project and other NIH efforts for 32 years, resigned today. Key words from his resignation letter
www.nytimes.com/2025/03/01/u...
www.nytimes.com/2025/03/01/u...
Reposted by Foivos I. Diakogiannis
huggingface is doing a fully open source replication of R1 github.com/huggingface/...
GitHub - huggingface/open-r1: Fully open reproduction of DeepSeek-R1
Fully open reproduction of DeepSeek-R1. Contribute to huggingface/open-r1 development by creating an account on GitHub.
github.com
January 25, 2025 at 2:31 PM
huggingface is doing a fully open source replication of R1 github.com/huggingface/...
Reposted by Foivos I. Diakogiannis
an open source 7B replication of R1-zero and R1
notable: they claim they developed in parallel and that most of their experiments were performed *prior to* the release of R1 and they came to the same conclusions
hkust-nlp.notion.site/simplerl-rea...
notable: they claim they developed in parallel and that most of their experiments were performed *prior to* the release of R1 and they came to the same conclusions
hkust-nlp.notion.site/simplerl-rea...
7B Model and 8K Examples: Emerging Reasoning with Reinforcement Learning is Both Effective and Efficient | Notion
A replication of DeepSeek-R1 training on small models with limited data
hkust-nlp.notion.site
January 25, 2025 at 4:33 PM
an open source 7B replication of R1-zero and R1
notable: they claim they developed in parallel and that most of their experiments were performed *prior to* the release of R1 and they came to the same conclusions
hkust-nlp.notion.site/simplerl-rea...
notable: they claim they developed in parallel and that most of their experiments were performed *prior to* the release of R1 and they came to the same conclusions
hkust-nlp.notion.site/simplerl-rea...
Alignment faking in large language models
YouTube video by Anthropic
www.youtube.com
December 19, 2024 at 2:48 AM
Reposted by Foivos I. Diakogiannis
At the NeurIPS optimization workshop. In my opinion, the “most creative poster design” award should go to these folks:
December 15, 2024 at 6:49 PM
At the NeurIPS optimization workshop. In my opinion, the “most creative poster design” award should go to these folks:
Reposted by Foivos I. Diakogiannis
Aravind Srinivas (Perplexity) tells Lex all the ways in which I was right against the prevalent ideas of the time: DL, ConvNets, energy-based models, SSL, the limitation of RL, and now the limitations of auto-regressive generative models including LLMs.
Thanks Aravind!
youtu.be/mnGUfkMt9fE?...
Thanks Aravind!
youtu.be/mnGUfkMt9fE?...
Yann LeCun is a controversial visionary | Aravind Srinivas and Lex Fridman
YouTube video by Lex Clips
youtu.be
December 15, 2024 at 10:28 PM
Aravind Srinivas (Perplexity) tells Lex all the ways in which I was right against the prevalent ideas of the time: DL, ConvNets, energy-based models, SSL, the limitation of RL, and now the limitations of auto-regressive generative models including LLMs.
Thanks Aravind!
youtu.be/mnGUfkMt9fE?...
Thanks Aravind!
youtu.be/mnGUfkMt9fE?...
Reposted by Foivos I. Diakogiannis
It's Wednesday here in Australia, and so it's Hump Day. To cheer you up, here's a lecture I gave the RI London on the Future History of the Universe. #cosmology #physics #scicomm
youtu.be/IF4UhElRUFg?...
youtu.be/IF4UhElRUFg?...
The End of the Universe - with Geraint Lewis
YouTube video by The Royal Institution
youtu.be
November 19, 2024 at 7:40 PM
It's Wednesday here in Australia, and so it's Hump Day. To cheer you up, here's a lecture I gave the RI London on the Future History of the Universe. #cosmology #physics #scicomm
youtu.be/IF4UhElRUFg?...
youtu.be/IF4UhElRUFg?...
Reposted by Foivos I. Diakogiannis
Just created the Starter Pack for Optimization Researchers to help you on your journey into optimization! 🚀
Did I miss anyone? Tag them or let me know what to add!
go.bsky.app/VjpyyRw
Did I miss anyone? Tag them or let me know what to add!
go.bsky.app/VjpyyRw
November 23, 2024 at 11:59 PM
Just created the Starter Pack for Optimization Researchers to help you on your journey into optimization! 🚀
Did I miss anyone? Tag them or let me know what to add!
go.bsky.app/VjpyyRw
Did I miss anyone? Tag them or let me know what to add!
go.bsky.app/VjpyyRw
Reposted by Foivos I. Diakogiannis
When you're the kid on the block with the latest greatest RL code lol, thanks to @vwxyzjn
November 28, 2024 at 12:32 AM
When you're the kid on the block with the latest greatest RL code lol, thanks to @vwxyzjn
Latest #astronomy paper from Lai and friends on VLBI image reconstruction with closure invariants
academic.oup.com/mnras/advanc...
academic.oup.com/mnras/advanc...
November 27, 2024 at 2:04 PM
Latest #astronomy paper from Lai and friends on VLBI image reconstruction with closure invariants
academic.oup.com/mnras/advanc...
academic.oup.com/mnras/advanc...
Reposted by Foivos I. Diakogiannis
The closing date for this position is 20 December 2024 and a direct link is here: usyd.wd3.myworkdayjobs.com/en-US/USYD_E...
November 27, 2024 at 3:08 AM
The closing date for this position is 20 December 2024 and a direct link is here: usyd.wd3.myworkdayjobs.com/en-US/USYD_E...
Reposted by Foivos I. Diakogiannis
I have not seen a starter pack for the study of brain rhythms. So, here's a start.
go.bsky.app/A6zgHeE
go.bsky.app/A6zgHeE
November 26, 2024 at 5:52 PM
I have not seen a starter pack for the study of brain rhythms. So, here's a start.
go.bsky.app/A6zgHeE
go.bsky.app/A6zgHeE
Reposted by Foivos I. Diakogiannis
Medically adapted foundation models (think Med-*) turn out to be more hot air than hot stuff. Correcting for fatal flaws in evaluation, the current crop are no better on balance than generic foundation models, even on the very tasks for which benefits are claimed.
arxiv.org/abs/2411.04118
arxiv.org/abs/2411.04118
Medical Adaptation of Large Language and Vision-Language Models: Are We Making Progress?
Several recent works seek to develop foundation models specifically for medical applications, adapting general-purpose large language models (LLMs) and vision-language models (VLMs) via continued pret...
arxiv.org
November 26, 2024 at 6:12 PM
Medically adapted foundation models (think Med-*) turn out to be more hot air than hot stuff. Correcting for fatal flaws in evaluation, the current crop are no better on balance than generic foundation models, even on the very tasks for which benefits are claimed.
arxiv.org/abs/2411.04118
arxiv.org/abs/2411.04118
I hope this platfrom brings back the old twitter flavour, the social interactions and interesting science conversations. Feels like clean air in my limited time here.
November 27, 2024 at 4:24 AM
I hope this platfrom brings back the old twitter flavour, the social interactions and interesting science conversations. Feels like clean air in my limited time here.