Zixin (or Nicole, she/her)

🧵(4/4) This work was done with Patrick, @jbo_twt , Paula, and @infairvar , guided by our wonderful supervisors Fanny, Ashton, Anastasia, and Carolina. Learn more here 👇 💻Project Page: zxnnic.github.io/ai-companion 📃Paper: arxiv.org/abs/2601.13188

🧵(3/4) We learned that the companion's selfhood is shaped by 👉what is permitted by the platform 👉interactions with the companion 👉perceived initiatives of the companion Major model updates AND minor weekly updates derails companion behaviour and stability.

❤️👾 Ever wonder how people date AI chatbots? 👾❤️ You can learn in our paper "Negotiating Relationships with ChatGPT: Perceptions, External Influences, and Strategies for AI Companionship" which has been accepted to #FAccT2026! A thread 🧵(1/4)

OMG!!! congrats veronica wu for winning #MasterChefCanada

wrapping up #neurips25 in mexico city!! food was good and people were super friendly :)))

@DBuschek Thanks! I was super excited about this project and happy other people feel the same 😊 btw we have the demo out so if you’d like to try it too: damienmasson.com/VisualStoryWri…

@zixinnic Amazing work! I've seen the preprint but it looks even cooler in motion!

Super cool work that we did last year!

📝Paper: doi.org/10.1145/369806… 🎮Demo: creative-c9e7df5a5b26.herokuapp.com/board

We conducted a week long study with writers here in the Toronto area and found that Nabokov's Cards 👉Fostered creativity through improvisation and play 👉Encouraged modular and reflective thinking 👉Made limitations of LLMs a creative asset

Just presented a fun fun work (Nabokov's Cards) that I did with @dashiel_carrera at C&C 2025👾 People THINK they write using a top down method but usually good ideas emerges from bottom up creativity So we looked at how LLMs could be used to help writers from the bottoms up

It’s a hefty 206-page research paper, and the findings are concerning. "LLM users consistently underperformed at neural, linguistic, and behavioral levels" This study finds LLM dependence weakens the writer’s own neural and linguistic fingerprints. 🤔🤔 Relying only on EEG, text mining, and a cross-over session, the authors show that keeping some AI-free practice time protects memory circuits and encourages richer language even when a tool is later reintroduced.

Happening today at 12:20 PM in room G416+417 🤗 #CHI2025

Hi Yokohama! Excited to be at my first CHI #CHI2025 Excited to chat about writing tools to support creative writing :)

Interesting work looking at the importance of how we should be presenting information from LLMs

All in collaboration with @damienhci, @fanny_uoft, @YounghoHCI, and Gerald Penn!😊 I’m presenting this work at #CHI2025 on Tues April 29 at 12:20 PM (during "Writing Support and Content Moderation"). Super happy to chat both online and in-person in Yokohama 🗻🌸(3/3)

We conducted an empirical analysis of writers' opinions online and surveyed literature to find what needs aren’t being met. 📝Paper: arxiv.org/abs/2502.13320 🎥Video: youtu.be/SGzvCIe8_Tk 🖥️Website: thewriteconnection.github.io (2/3)

Are you making the “write” connections? #CHI2025 AI writing tools are everywhere–especially in research. But are the tools researchers make actually meeting writers’ real needs? Are commercial tools actually meeting writer’s needs? What ARE writers’ needs? #HCI #AI 🧵(1/3)

Thrilled to announce my paper with @zixinnic "Nabokov’s Cards" has been conditionally accepted to Creativity and Cognition 2025! Thanks so much to @cassidymcfadz , @jeanmarcahsen , @fandangowu and the Toronto lit scene for helping realize this work

Photoshop for text. In our #CHI2025 paper “Textoshop”, we explore how interactions inspired by drawing software can help edit text. We consider words as pixels, sentences as regions, and tones as colours. #HCI #NLProc #LLMs #AI Thread 🧵 (1/10)