Andrea Panizza
@andreapanizza.bsky.social
ML, trekking, enjoying life
Reposted by Andrea Panizza
Reposted by Andrea Panizza
Interviewer: Can you explain this gap in your resume?
LM researcher: You're right to wonder about the gaps in my resume! They are more common than people think, and there are many valid reasons why someone might have them. Here are some of the most frequent reasons you might see a gap:
LM researcher: You're right to wonder about the gaps in my resume! They are more common than people think, and there are many valid reasons why someone might have them. Here are some of the most frequent reasons you might see a gap:
Interviewer: Can you explain this gap in your resume?
Egyptologist: why yes, that’s the Fourth Intermediate Period, when I labored without Ma’at…
Egyptologist: why yes, that’s the Fourth Intermediate Period, when I labored without Ma’at…
Interviewer: Can you explain this gap in your resume?
Rare book cataloguer: [1] i, from 1-9 (8), took a break at [2] I-III8, IV10, then was Cited In and had to Bound-with 2°: πA⁶(πA1+1, πA5+1.2), A-2B6, 2C2, x4, “gg3.4″(±”gg3″), ¶-2¶6, 3¶1, 2a- 2f6, 2g2, “Gg6“, 2h6, 2k-3b7. But eventually, [n.d.]
Rare book cataloguer: [1] i, from 1-9 (8), took a break at [2] I-III8, IV10, then was Cited In and had to Bound-with 2°: πA⁶(πA1+1, πA5+1.2), A-2B6, 2C2, x4, “gg3.4″(±”gg3″), ¶-2¶6, 3¶1, 2a- 2f6, 2g2, “Gg6“, 2h6, 2k-3b7. But eventually, [n.d.]
May 12, 2025 at 11:39 PM
Interviewer: Can you explain this gap in your resume?
LM researcher: You're right to wonder about the gaps in my resume! They are more common than people think, and there are many valid reasons why someone might have them. Here are some of the most frequent reasons you might see a gap:
LM researcher: You're right to wonder about the gaps in my resume! They are more common than people think, and there are many valid reasons why someone might have them. Here are some of the most frequent reasons you might see a gap:
Reposted by Andrea Panizza
Finally found the time to watch it in full: one of the most interesting and thought provoking LLM conference I’ve seen in a while. You see LLM as latent attention graphs and many structural oddities suddenly falls into place. www.youtube.com/watch?v=J1YC...
April 12, 2025 at 6:45 AM
Finally found the time to watch it in full: one of the most interesting and thought provoking LLM conference I’ve seen in a while. You see LLM as latent attention graphs and many structural oddities suddenly falls into place. www.youtube.com/watch?v=J1YC...
Reposted by Andrea Panizza
Parents Gently Explain To Child That Their Money In Heaven Now
Parents Gently Explain To Child That Their Money In Heaven Now
HUNTSVILLE, AL—In an effort to comfort the child by telling her the funds had gone to a far better place, local parents Blake and Allison McKee gently explained to their daughter Friday that their mon...
theonion.com
April 4, 2025 at 9:30 PM
Parents Gently Explain To Child That Their Money In Heaven Now
Reposted by Andrea Panizza
More than 250 people have already enrolled in the Causal Secrets Mini-Course!
All but one review so far are 5-star.
It's free for everyone!
Share it with your friend!
https://bit.ly/4ic4VK4
#CausalSky
All but one review so far are 5-star.
It's free for everyone!
Share it with your friend!
https://bit.ly/4ic4VK4
#CausalSky
March 11, 2025 at 9:29 AM
More than 250 people have already enrolled in the Causal Secrets Mini-Course!
All but one review so far are 5-star.
It's free for everyone!
Share it with your friend!
https://bit.ly/4ic4VK4
#CausalSky
All but one review so far are 5-star.
It's free for everyone!
Share it with your friend!
https://bit.ly/4ic4VK4
#CausalSky
Reposted by Andrea Panizza
One of the first papers I've seen with RLVR / reinforcement finetuning of vision language models
Looks about as simple as we would expect it to be, lots of details to uncover.
Liu et al. Visual-RFT: Visual Reinforcement Fine-Tuning
buff.ly/DbGuYve
(posted a week ago, oops)
Looks about as simple as we would expect it to be, lots of details to uncover.
Liu et al. Visual-RFT: Visual Reinforcement Fine-Tuning
buff.ly/DbGuYve
(posted a week ago, oops)
March 10, 2025 at 3:44 PM
One of the first papers I've seen with RLVR / reinforcement finetuning of vision language models
Looks about as simple as we would expect it to be, lots of details to uncover.
Liu et al. Visual-RFT: Visual Reinforcement Fine-Tuning
buff.ly/DbGuYve
(posted a week ago, oops)
Looks about as simple as we would expect it to be, lots of details to uncover.
Liu et al. Visual-RFT: Visual Reinforcement Fine-Tuning
buff.ly/DbGuYve
(posted a week ago, oops)
Reposted by Andrea Panizza
Here's the handout for my "Cutting-edge web scraping techniques" workshop at #NICAR2025 this morning github.com/simonw/nicar...
Plus some extra notes on the custom software I built to support the workshop: simonwillison.net/2025/Mar/8/c...
Plus some extra notes on the custom software I built to support the workshop: simonwillison.net/2025/Mar/8/c...
Cutting-edge web scraping techniques at NICAR
Here's the handout for a workshop I presented this morning at [NICAR 2025](https://www.ire.org/training/conferences/nicar-2025/) on web scraping, focusing on lesser know tips and tricks that became po...
simonwillison.net
March 8, 2025 at 7:29 PM
Here's the handout for my "Cutting-edge web scraping techniques" workshop at #NICAR2025 this morning github.com/simonw/nicar...
Plus some extra notes on the custom software I built to support the workshop: simonwillison.net/2025/Mar/8/c...
Plus some extra notes on the custom software I built to support the workshop: simonwillison.net/2025/Mar/8/c...
Reposted by Andrea Panizza
This thing now deserves its own name
March 6, 2025 at 9:04 PM
This thing now deserves its own name
Reposted by Andrea Panizza
Reposted by Andrea Panizza
I already advertised for this document when I posted it on arXiv, and later when it was published.
This week, with the agreement of the publisher, I uploaded the published version on arXiv.
Less typos, more references and additional sections including PAC-Bayes Bernstein.
arxiv.org/abs/2110.11216
This week, with the agreement of the publisher, I uploaded the published version on arXiv.
Less typos, more references and additional sections including PAC-Bayes Bernstein.
arxiv.org/abs/2110.11216
March 5, 2025 at 1:16 AM
I already advertised for this document when I posted it on arXiv, and later when it was published.
This week, with the agreement of the publisher, I uploaded the published version on arXiv.
Less typos, more references and additional sections including PAC-Bayes Bernstein.
arxiv.org/abs/2110.11216
This week, with the agreement of the publisher, I uploaded the published version on arXiv.
Less typos, more references and additional sections including PAC-Bayes Bernstein.
arxiv.org/abs/2110.11216
Reposted by Andrea Panizza
My self driving car writeup from December (needs an update) open.substack.com/pub/itcanthi...
Self Driving Cars are At A Transition Point
Cruise leaves the game as Waymo and Tesla ramp up
open.substack.com
March 5, 2025 at 5:25 AM
My self driving car writeup from December (needs an update) open.substack.com/pub/itcanthi...
Reposted by Andrea Panizza
This is revised down to -2.8%, and partly precipitated the big flush today. Probably some sovereign wealth fund exited positions across Nasdaq and S&P given the volume of sales.
March 3, 2025 at 9:40 PM
This is revised down to -2.8%, and partly precipitated the big flush today. Probably some sovereign wealth fund exited positions across Nasdaq and S&P given the volume of sales.
Reposted by Andrea Panizza
If you look at most of the models we've received from OpenAI, Anthropic, and Google in the last 18 months you'll hear a lot of "Most of the improvements were in the post-training phase."
Here's a simple analogy for how so many gains can be made on mostly the same base model:
Here's a simple analogy for how so many gains can be made on mostly the same base model:
March 3, 2025 at 4:22 PM
If you look at most of the models we've received from OpenAI, Anthropic, and Google in the last 18 months you'll hear a lot of "Most of the improvements were in the post-training phase."
Here's a simple analogy for how so many gains can be made on mostly the same base model:
Here's a simple analogy for how so many gains can be made on mostly the same base model:
Reposted by Andrea Panizza
More evidence of the importance of training analysis for interp! Induction heads might serve as *preliminary* function vector heads (which directly compute in-context learning tasks). Ultimately, LMs rely on FV heads more than IH heads for ICL. from @kayoyin.bsky.social
Which Attention Heads Matter for In-Context Learning?
Large language models (LLMs) exhibit impressive in-context learning (ICL) capability, enabling them to perform new tasks using only a few demonstrations in the prompt. Two different mechanisms have be...
arxiv.org
March 3, 2025 at 4:51 PM
More evidence of the importance of training analysis for interp! Induction heads might serve as *preliminary* function vector heads (which directly compute in-context learning tasks). Ultimately, LMs rely on FV heads more than IH heads for ICL. from @kayoyin.bsky.social
Reposted by Andrea Panizza
Impressive piece of work by Soumya Mukherjee and Bharath Sriperumbudur: arxiv.org/abs/2502.20755
Minimal optimal kernel two-sample tests with random Fourier features.
Minimal optimal kernel two-sample tests with random Fourier features.
Minimax Optimal Kernel Two-Sample Tests with Random Features
Reproducing Kernel Hilbert Space (RKHS) embedding of probability distributions has proved to be an effective approach, via MMD (maximum mean discrepancy) for nonparametric hypothesis testing problems ...
arxiv.org
March 3, 2025 at 7:05 AM
Impressive piece of work by Soumya Mukherjee and Bharath Sriperumbudur: arxiv.org/abs/2502.20755
Minimal optimal kernel two-sample tests with random Fourier features.
Minimal optimal kernel two-sample tests with random Fourier features.
Reposted by Andrea Panizza
Our Workshop on Uncertainty Quantification for Computer Vision goes to @cvprconference.bsky.social this year!
We have a super line-up of speakers and a call for papers.
This is a chance for your paper to shine at #CVPR2025
⏲️ Submission deadline: 14 March
💻 Page: uncertainty-cv.github.io/2025/
We have a super line-up of speakers and a call for papers.
This is a chance for your paper to shine at #CVPR2025
⏲️ Submission deadline: 14 March
💻 Page: uncertainty-cv.github.io/2025/
February 28, 2025 at 7:28 AM
Our Workshop on Uncertainty Quantification for Computer Vision goes to @cvprconference.bsky.social this year!
We have a super line-up of speakers and a call for papers.
This is a chance for your paper to shine at #CVPR2025
⏲️ Submission deadline: 14 March
💻 Page: uncertainty-cv.github.io/2025/
We have a super line-up of speakers and a call for papers.
This is a chance for your paper to shine at #CVPR2025
⏲️ Submission deadline: 14 March
💻 Page: uncertainty-cv.github.io/2025/
Reposted by Andrea Panizza
A new paper by Vovk that continues exploring properties of so-called "randomness predictors" (compared to "conformal predictors").
www.arxiv.org/abs/2502.19254
www.arxiv.org/abs/2502.19254
February 27, 2025 at 11:25 PM
A new paper by Vovk that continues exploring properties of so-called "randomness predictors" (compared to "conformal predictors").
www.arxiv.org/abs/2502.19254
www.arxiv.org/abs/2502.19254
Reposted by Andrea Panizza
I am happy to announce that the Kakeya set conjecture, one of the most sought after open problems in geometric measure theory, has now been proven (in three dimensions) by Hong Wang and Joshua Zahl! arxiv.org/abs/2502.17655 I discuss some ideas of the proof at terrytao.wordpress.com/2025/02/25/t...
Volume estimates for unions of convex sets, and the Kakeya set conjecture in three dimensions
We study sets of $δ$ tubes in $\mathbb{R}^3$, with the property that not too many tubes can be contained inside a common convex set $V$. We show that the union of tubes from such a set must have almos...
arxiv.org
February 26, 2025 at 4:49 AM
I am happy to announce that the Kakeya set conjecture, one of the most sought after open problems in geometric measure theory, has now been proven (in three dimensions) by Hong Wang and Joshua Zahl! arxiv.org/abs/2502.17655 I discuss some ideas of the proof at terrytao.wordpress.com/2025/02/25/t...
Reposted by Andrea Panizza
A distributionally robust extension of conformal prediction arxiv.org/abs/2502.14105
At a high level the formulation is straightforward:
At a high level the formulation is straightforward:
February 25, 2025 at 8:50 AM
A distributionally robust extension of conformal prediction arxiv.org/abs/2502.14105
At a high level the formulation is straightforward:
At a high level the formulation is straightforward:
Reposted by Andrea Panizza
New YouTube video posted
"Measuring the Earth...from a vacation photo!"
(correct link this time: youtu.be/038AkmPvltA)
"Measuring the Earth...from a vacation photo!"
(correct link this time: youtu.be/038AkmPvltA)
February 22, 2025 at 4:17 PM
New YouTube video posted
"Measuring the Earth...from a vacation photo!"
(correct link this time: youtu.be/038AkmPvltA)
"Measuring the Earth...from a vacation photo!"
(correct link this time: youtu.be/038AkmPvltA)
Reposted by Andrea Panizza
Star of Leopard News Asks That Friend’s Face be Spared
Jesse Watters Makes On-Air Plea to Trump for Veteran Friend Who Got ‘DOGE’d’ In Pentagon Cuts
Watters made an on-air plea for a military veteran friend was “DOGE’d” in Elon Musk’s Pentagon cuts.
www.mediaite.com
February 20, 2025 at 12:09 PM
Star of Leopard News Asks That Friend’s Face be Spared