Are you a US ugrad student interested in doing summer research in HCI? We've just opened up applications for our DUB Undergraduate Summer REU program! Interns will be matched with a HCI & Design at UW faculty mentor. ~9 weeks, paid, on UW campus. Deadline is 2/2: dub.washington.edu/posts/2024/dub…

You can add alt texts to figures in LaTeX with `alt={...}`. The alt texts won't necessarily appear in the generated PDF until new versions of LaTeX but let's start with this good practice now.

I've had fun playing with Observable Framework to deploy dashboards and web apps. Here's an example site integrating Interactive Data Lab Mosaic and @DuckDB for scalable visualization: uwdata.github.io/mosaic-framewo…

Do you make visualizations with Vega-lite, Altair, or Plot.ly? Have you have used their example galleries? If so please participate in our (paid!) study! (30min/15$) We want to understand how/why folks use galleries To sign up, fill out the survey linked below!

If you’re excited about Observable Framework’s new SQL cells, wait till you see our Mosaic integration! The code: --- sql: gaia: ./gaia-sample.parquet --- # Hello, vgplot ```js echo vg.plot(vg.raster(vg.from("gaia"), {x: "ra", y: "dec", fill: "density"})) ```

I've always been excited about the hybrid client/server execution model in MotherDuck but seeing how well the WebAssembly client works with Interactive Data Lab Mosaic is even more exciting. In the video below I am looking at 10m data points in my browser.

Observable Framework 1.3 🆕 integrates Interactive Data Lab’s Mosaic vgplot, which can concisely expressive performant coordinated views of millions of data points. observablehq.com/framework/lib/…

Do you use shell scripts? 🐚 📜 I'm conducting a study of how new tools might help. I'd love to talk with you about your shell experiences, and to run our new prototype by you. Interested? Heck yeah you are, here's a form: docs.google.com/forms/d/e/1FAI… (Novices & experts plz!)

“Can we get a new text analysis tool?” “No—we have Topic Model at home” Topic Model at home: outputs vague keywords; needs constant parameter fiddling🫠 Is there a better way? We introduce LLooM, a concept induction tool to explore text data in terms of interpretable concepts🧵

Excited to analyze text at the level of *interpretable concepts*, addressing many weaknesses of topic models we’ve found in the past - and with more control & revision by analysts. Also a fun collaboration between Interactive Data Lab and Stanford Human-Computer Interaction Group, led by the impressive Michelle Lam!

Frustrated with topic models? Wish emergent concepts were interpretable, steerable, and able to classify new data? Check out our #CHI2024 talk on Tues 9:45am in 316C (Politics of Datasets)! Or try LLooM, our open-sourced tool :) ✨ github.com/michelle123lam…

I'm super happy that Hex donated VegaFusion to the Vega org to bring it together with Altair which is also now under the Vega org. It's been amazing to collaborate with Jon Mease and I look forward to continue doing so.

Do you use shell scripts? 🐚 📜 I'm conducting a study of how a new tool might help. I'd love to have you try it out! Interested? I knew it! Here's a form: forms.gle/ZRaCQpucB5KnTS… Novices & experts plz! 🙏 P.S. Yeah I posted a similar thing in March; this is Part 2! Woohoo!

🚀 New Benchmark Alert: ⚙️ BLADE — A benchmark with expert-annotated analyses, evaluating how #LLM agents tackle open-ended, data-driven scientific research questions. Results show top models reach only ~44% F1. 📄 arxiv.org/abs/2408.09667 👨🏻‍💻 github.com/behavioral-dat… 🧵1/9