Samuel G. Balter, Ethan Jerzak, Connor T. Jerzak: Multiplication in Multimodal LLMs: Computation with Text, Image, and Audio Inputs https://arxiv.org/abs/2604.18203 https://arxiv.org/pdf/2604.18203 https://arxiv.org/html/2604.18203
Thanks, Ingo! Graham Goff and I have put a lot of work into this package. {DAGassist} can estimate a wide range of models and classify covariate roles after you make your DAG either with {ggdag} or {daggitty}. We hope that the package can make DAG-consistent estimation a little easier for everyone!
UT Austin PolMeth Speaker Series: Grad Student Lightning Talks this week! ⚡️
🗓️ Thurs, Apr 16 @ 12:30pm
📍 Batt 5.108 (Lunch provided!)
🗣️ Rachel Jeon: Activism & Treaty Design
🗣️ Sophia Lander: Tech Innovation in PPPs
🗣️ Allison Verrilli: Policing Federalism & Local Safety
We're hiring a Research Support Associate at MIT Political Science. Work with me, In Song Kim, Rich Nielsen, and Naoki Egami on methods and political economy. Great opportunity for those considering a PhD in the social sciences. One-year position with possible extension.
Excited to welcome Gengchen Mai to the UT Austin Methods Speaker Series this Thursday, April 2!
Talk: “Spatial Representation Learning: What, How, and Why” on latest GeoAI & scientific applications.
🕛 12:00-1:30pm
📍 BAT 5.108 (5th floor Batts Hall), UT Austin campus
Join us!
New grad course at UT Austin this coming Fall (2026); look for this in the catalog -- "Gov 385L: Causal ML & EO for Social Sci"
Cool workshop upcoming this May -- join if you can.
"We take rationality seriously, but should we? Why is a rational agent better in some respect than an irrational one? Is rationality just instrumentally useful, or does it have intrinsic value?"
ejerzak.github.io
Join us on Thursday, 29 January, at *13:30 CET for the International Roundtable on Computational Social Science with
Bedoor AlShebli (NYU Abu Dhabi)
🔹 The Science of Science... and its Human Side 🔹 More info: liu.se/en/event/int...
FastRerandomize site: fastrerandomize.github.io
Paper (forthcoming in SoftwareX): arxiv.org/abs/2501.07642
FastRerandomize unites batched candidate generation, key-only storage & design-respecting inference, while using the SoTA in accelerated computing (T/GPU). That design delivers order‑of‑mag speedups & enables tighter covariate balance in high-dim experiments⇢more precise causal estimates, lower cost
The remains of Gresham castle photographed with a regular camera and how it looks with a LiDar drone camera.
Calling those looking to get feedback on their quant work📢: Join us at @TexasAM for TexMeth2026 this Feb. Great opportunity to connect with the regional community and showcase your research. The sun will still be strong for those coming from the north!🤝📊☀️
Apply: docs.google.com/forms/d/e/1F...
Application information: liberalarts.utexas.edu/government/g...
More about Austin: connorjerzak.com/ut-austin/
Application deadline for UT Austin Gov - PhD admissions is 12/10 this cycle. More information below. I am hearing that the Department will not be shrinking its cohort this admission cycle!
The 8th European Conference on Social Networks (EUSN 2026) will take place on 11-15 August 2026 in Norrköping, Sweden, hosted by @iasliu.bsky.social
⌛ Deadline for workshop and session proposals: 1 December 2025
More information: liu.se/en/event/eus...
Looking forward to giving a presentation on the planetary causal inference paradigm at Department of Methodology, LSE, today. Join us if you are around.
www.lse.ac.uk/methodology/...
poster that reads 'seminar series, 21 nov 2-3pm, professor adel daoud
Sign up for this week's Seminar Series with @adeldaoud.bsky.social
💡 21 Nov, 2 - 3pm
🌎Planetary Causal Inference: combining computer vision and earth observation to analyse disparities in health and living conditions among neighbourhoods in Africa, 1990 to date
www.eventbrite.com/e/department...
If you are in the vicinity of Oxford University, join me today at 4 pm for my talk at Nuffield College. Title: "Planetary Causal Inference: combining computer vision and earth observation to..."
Link to event details: [Sociology Seminar at Nuffield College](www.nuffield.ox.ac.uk/news-events/...)
Markus B. Pettersson, Adel Daoud: Leveraging Compact Satellite Embeddings and Graph Neural Networks for Large-Scale Poverty Mapping https://arxiv.org/abs/2511.01408 https://arxiv.org/pdf/2511.01408 https://arxiv.org/html/2511.01408
"Counterfactuals. At planetary scale."
The Planetary Causal Inference (PCI) book launches soon:
planetarycausalinference.org/book-launch/
Excited to see our PhD student Andrés Cruz (@arcruz0) presenting his work "Sensitivity Bounds for Uncertainty Propagation" at LAPolMeth IX in Viña del Mar, Chile.
Catch it during the poster session on Nov 22 (12:45-14:00, Building C). Don't miss this work on uncertainty in quant social science!
Debiasing ML predictions for causal inference—no new labels needed. We propose Tweedie’s correction to fix shrinkage enabling “one map, many trials.”
arxiv.org/abs/2508.01341
#CausalInference #MachineLearning #EarthObservation #PovertyMapping
Will definitely post a public version to the webpage below once final details are ironed out! Will shoot you a message once its live 🙏
connorjerzak.com/ai-programmi...
🥳 Join us in celebrating @alongcamejones.bsky.social and his legacy with the Policy Agendas Project. Excited to see what @seanmtheriault.bsky.social and Derek Epp do next! 🤟
ML wealth maps from space 🛰️ are great, but suffer from "shrinkage" bias, which waters down policy impact results (causal inference). We developed correction methods that fix this bias *without* new data.
arxiv.org/abs/2508.01341
#CausalInference #DataForGood #AI #PovertyMapping #EarthObservation
Are you interested in using earth observation data and deep learning for estimating poverty in Africa, poverty specifically? Then apply to this research engineer position at AI and Global Development Lab. More info about our research at www.aidevlab.org . Deadline is 7th Nov.
lnkd.in/ePX7VT-K
New undergrad course (UT Austin, this spring): Gov 355M - AI & Programming Foundations for Computational Social Science.
We'll treat curiosity as our raw material and AI+computation as our foundry - learning how the frontiers of technology can accelerate discovery in the social sciences + beyond.
A little less than 2 weeks to submit an abstract proposal for PolMeth - Europe. Join us in Dublin this coming May to share your new political science research using cutting edge methods!!
Check out the data paper (open access) and the publicly available data: We hope that many people will use it in their research and for other purposes.