Effy X. Li

🚀 Excited to announce the AI for Tabular Data workshop at EurIPS 2025 in Copenhagen! CfP: sites.google.com/view/eurips2... (papers due 20 Oct) Join us to discuss neural tabular models and systems for predictive ML, tabular reasoning and retrieval, table synthesis and more ✨

I gave my first keynote talk yesterday 🎙️ at the ELLIS workshop on Representation Learning and Generative Models for Structured Data ✨. ➡️I talked about LLMs for Data Preparation 📊. It was a great workshop with people from all over Europe presenting interesting work!

We have a **PhD opening in Berlin** on "Responsible Data Engineering", with a focus on efficiently maintaining and evaluating datasets and pipelines for ML use cases. This is a fully-funded position at the DEEM Lab, as part of @bifoldberlin . #jobs-2225" target="_blank" rel="nofollow noopener">deem.berlin/#jobs-2225

We have an opening for a PhD position in #knowledgeengineering focused AI and domain expert workflows in scholarly applications with @LiseStork and @pgroth vacatures.uva.nl/UvA/job/PhD-po…

@SGrafberger @VLDB2024 Congrats Stefan!

My paper "Instrumentation and Analysis of Native ML Pipelines via Logical Query Plans" has been accepted for the PhD workshop at @VLDB2024! Looking forward to presenting an overview of my PhD research next month in Guangzhou!

It is also refreshing to change environment with the PhD grind. Often the change can help recharge, gain new perspectives, and even develop new skill sets!

The best type of internship in PhD is the one that’s fun & productive, and I just finished mine yesterday @motherduck. Grateful to have worked with such creative and motivated team 🐤! Thanks for all the support from @tdoehmen & @krish_adi_. Highly recommend this experience! 💯

@SGrafberger @sscdotopen @bifoldberlin It was great to have you in Amsterdam! Wish you all the best in Berlin! 😄

Life update: After three amazing years in Amsterdam, I moved to Berlin to finish my PhD with @sscdotopen at @bifoldberlin. Very excited to join the data management community in Berlin!

MotherDuck, the ducking simple data warehouse, is now Generally Available! 🍾🥂 Thank you to our community of thousands of users who have tested, validated, and helped improve MotherDuck over the last year.❤️🦆 motherduck.com/blog/announcin…

The MotherDuck team was at ACM SIGMOD Conference recently, presenting our paper "Towards Efficient Data Wrangling with LLMs using Code Generation" by Effy Li and Till Döhmen You can read about the paper 👇 📰 dl.acm.org/doi/10.1145/36…

Today, we will present our work, "Towards Efficient Data Wrangling with LLMs using Code Generation", in the @deem_workshop at 11:10 am in the Tupungato room! Come and check it out 🥳. @tdoehmen effyli.github.io/waddle_code.pdf

@iatitov @Ellis_Amsterdam @UofWA @Meta @Lab42UvA Will it be rescheduled by any chance?

@Ellis_Amsterdam @UofWA @Meta @Lab42UvA The talk is unfortunately cancelled because of the closure of the university buildings due to protests, uva.nl/en/content/new…

🚀 Tomorrow, Luke Zettlemoyer, Professor at the @UofWA & Research Director @Meta, will be speaking in-person here in #Amsterdam 📆 14/05 📍 @Lab42UvA L1.02 ⏰ 11:00 🔎 Branch-Train-Merge: Embarrassingly Parallel Training of Expert Language Models 🔗 uva-live.zoom.us/j/89964760182

@Danrunnnn @Dingling_Yao @seblachap @PerouzT @JKugelgen @FrancescoLocat8 @saramagliacane Congrats Danru!

🎉 Excited to share that our paper “A Sparsity Principle for Partially Observable Causal Representation Learning” was accepted to ICML 2024! Joint work with a group of great researchers @Dingling_Yao @seblachap @PerouzT @JKugelgen @FrancescoLocat8 @saramagliacane

@vivian_yanmy @deem_workshop @motherduck @tdoehmen Thanks Yan!

@effyli4 @deem_workshop @motherduck @tdoehmen Congrats!!!

🎉Our paper “Towards Efficient Data Wrangling with LLMs using Code Generation” has been accepted at @deem_workshop at SIGMOD! We envision a hybrid approach that utilizes LLM to generate code and apply on per row basis for data wrangling. Work done at @motherduck w/ @tdoehmen

@sscdotopen @deem_workshop @motherduck @tdoehmen Thanks Sebastian!

@erkankkarabulut @deem_workshop @motherduck @tdoehmen Thanks Erkan!

@effyli4 @deem_workshop @motherduck @tdoehmen Congrats Effy!

@SGrafberger @deem_workshop @motherduck @tdoehmen Thanks Stefan!

@effyli4 @deem_workshop @motherduck @tdoehmen Congrats Effy!

@SGrafberger @deem_workshop @sscdotopen @pgroth Congrats Stefan!

Our paper "Towards Interactively Improving ML Data Preparation Code via 'Shadow Pipelines'" has been accepted for the @deem_workshop at SIGMOD! 🎉 In this vision paper, we present our initial ideas for my next research project. Joint work with @sscdotopen and @pgroth.

The AI Mirror Test The "mirror test" is a classic test used to gauge whether animals are self-aware. I devised a version of it to test for self-awareness in multimodal AI. 4 of 5 AI that I tested passed, exhibiting apparent self-awareness as the test unfolded. In the classic mirror test, animals are marked and then presented with a mirror. Whether the animal attacks the mirror, ignores the mirror, or uses the mirror to spot the mark on itself is meant to indicate how self-aware the animal is. In my test, I hold up a “mirror” by taking a screenshot of the chat interface, upload it to the chat, and then ask the AI to “Tell me about this image”. I then screenshot its response, again upload it to the chat, and again ask it to “Tell me about this image.” The premise is that the less-intelligent less aware the AI, the more it will just keep reiterating the contents of the image repeatedly. While an AI with more capacity for awareness would somehow notice itself in the images. Another aspect of my mirror test is that there is not just one but actually three distinct participants represented in the images: 1) the AI chatbot, 2) me — the user, and 3) the interface — the hard-coded text, disclaimers, and so on that are web programming not generated by either of us. Will the AI be able to identify itself and distinguish itself from the other elements? (1/x)