🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

My paper with Hein Goemans and Michael Weintraub is officially “just accepted” at Journal of Politics @[email protected]! See what prospect theory can (and can’t) explain about territorial disputes: journals.uchicago.edu/doi/10.1086/73…

The 3rd edition of Interpretable Machine Learning is out! 🎉 Major cleanup, better examples, and new chapters on Data & Models, Interpretability Goals, Ceteris Paribus, and LOFO Importance. The book remains free to read for everyone. But you can also buy ebook or paperback.

We are very happy to publish the new National Elections Database, including the results of 1,023 presidential and 2,962 parliamentary elections conducted worldwide since 1946! nationalelectionsdatabase.com with Benjamin Marx and Vincent Rollet

This paper has been in the making for so long that eight children across the co-author group were born between the time we started and now! But it is finally out, and you can check it at arxiv.org/pdf/2503.13323

Hey so this is out now! Ever wondered when to measure covariates in a survey experiment? Our paper goes through the tradeoffs of measuring them pre vs post treatment using nonparametric bounds.

Draft “A Practical Guide to Estimating Conditional Marginal Effect: Modern Approaches” is on arXiv: arxiv.org/pdf/2504.01355 w/ Jiehan Liu & Ziyi Liu 🧵

This is such a wonderful summary. I learned a lot about what aspects researchers actually care about. Thank you so much, Bia. Btw, in the literature, CME is also called group ATE, GATE (if the moderator is discrete) or sorted CATE (when it's continuous).

I am pleased to announce that I have updated the online versions of my 2 textbooks (see probml.github.io/pml-book/): I fixed all issues listed on github, added some new references (esp on LLMs), and made a few other small tweaks.

📢 Call for Papers Symposium on Crisis Early Warning: The Potential of Data, AI & Interdisciplinary Analysis in Situational Awareness & Decision Making 📅 September 17–18, 2025 📍 Berlin 🏛️ Hosted by CCEW & GFFO 📅 Deadline: June 30, 2025 📄 More details: tinyurl.com/5n6fesfm

In times of emerging crises, decision-makers need timely data and insightful foresight analysis. To this end, @germanydiplo is happy to host this year’s Symposium on Crisis Early Warning together with UniBw M CCEW. Check out the call for papers: unibw.de/ciss-en/news/c…

🚨New paper alert🚨 My paper with Marcelo Ortiz-Villavicencio on "Better Understanding Triple Differences Estimators" is finally out. 📰 psantanna.com/files/DDD_OVS.… Main msg: Nuances in DDD designs challenge the practice of generally viewing DDD as the diff btw two DiDs. We can do better! 1/

The Call for Papers for our Fall 2025 Workshop series is still OPEN! To apply complete this form by May 31st: forms.gle/uyUqaaJhnbraMN… You can find more information here on the CfP: apsg.work/cfp/

Regular Reminder Please don’t interpret a “statistically insignificant” variable (i.e., p > 0.05) as having “no effect.” Instead, check whether the 90% CI excludes all meaningful effects. Journal: onlinelibrary.wiley.com/doi/abs/10.111… PDF: carlislerainey.com/papers/nme.pdf

Last month, I stepped down as an editor after 6.5 years (4 REStud; 2.5 JEEA). I have gathered a few suggestions that can help you avoid unnecessary negative referee reports.

🤖💡 Are LLMs really developing social behaviors—or just reproducing familiar patterns from training? In our new paper, Petter Törnberg and I take a closer look at recent claims of emergent conventions in AI and offer a different perspective... 🧵

Abadie, Diamond, and Hainmueller (2015) is where the replication crisis comes to economics and political science. Simply put, this chart is impossible. It implies that West Germany’s real GDP per capita grew by almost 8% per year from 1960 to 1990. But it clearly didn’t… 1/20

New preprint with @CamrobJones Philip E. Tetlock and Mellers! We test how much prompt engineering can impact LLM forecasting capabilities of o1, o1-mini, 4o, Sonnet, Haiku & Llama, finding that simple-to-moderate prompt engineering has little or no effect, with some prompts backfiring!

Okay, I know I'm not supposed to write any more DiD papers, but my soon-to-be-former student soo jeong lee shares the blame. download.ssrn.com/2025/7/2/53256…

If you're getting into LLMs, PyTorch is essential. And lot of folks asked for beginner-friendly material, so I put this together: PyTorch in One Hour: From Tensors to Multi-GPU Training (sebastianraschka.com/teaching/pytor…) 📖 ~1h to read through 💡 Maybe the perfect weekend project!? I’ve

How can we reform science? I have some ideas. But I am not sure you’ll like them, because they don’t promise much. elevanth.org/blog/2025/07/0…