@Freya_Wyy

PhD student @umdcs | previously MS @TAMU

cyberwenyi2357.github.io
Joined June 2021
  • Tweets39
  • Following260
  • Followers206
1Media
What are users thinking during their interactions with LLMs? We introduce ThoughtTrace — the first large-scale dataset that captures what users think during real-world human–AI conversations, not just what they type. → 10,174 thought annotations → 2,155 multi-turn conversations, 17,058 turns → 1,058 users → 20 LLMs These thoughts improve user behavior prediction (+41.7%) and model alignment (+25.6%). This opens a new paradigm of user-centric LLM research. Full information in the thread 🧶 Read our paper: arxiv.org/abs/2605.20087 Check our project website: thoughttrace-project.github.…
10
37
8
139
75,348
Yi Wen retweeted
Muhan Gao@muhan_gaoMay 26
🤖 We often talk about “context rot”: LLMs get worse as context grows. But once distracting information enters, is it just “a bit more noise → a bit worse performance”? Our #ICML2026 paper finds: no! 🤯 Instead, we reveal a striking "First Drop of Ink" effect: the first very few hard distractors do almost all of the damage, exactly like how one drop of ink clouding clear water. Paper link: arxiv.org/abs/2605.10828
6
15
4
48
10,900
Yi Wen retweeted
Anthropic
@AnthropicAIMay 7
New Anthropic research: Natural Language Autoencoders. Models like Claude talk in words but think in numbers. The numbers—called activations—encode Claude’s thoughts, but not in a language we can read. Here, we train Claude to translate its activations into human-readable text.
581
1,664
615
16,479
2,599,599
Yi Wen retweeted
ICLR@iclr_confApr 23
Announcing the #ICLR2026 Outstanding Paper Awards 🏆 Congratulations to:
7
98
9
1,015
87,335
Yi Wen retweeted
Meng XIA@Iriskie_XiaApr 8
📢 Call for Papers! Join our #AIED2026 workshop on Next-Gen Learning Interfaces (AI × HCI × Learning Sciences) 📍 South Korea 🔗 extendedhorizon.com/aied2026… Please share or submit! @HCI_Prof_YC @QiaoJin5 @yshi0526 @pauldenny @DrTiffanyBarnes @qingsongedu @vincentaleven
4
40
5,007
Yi Wen retweeted
Zora Wang@ZhiruoWFeb 11
‼️Position: AI coding agent research needs recalibration. We've heavily optimized for solo autonomy, and far less for designing agents that empower the humans using them. It’s time to build human-centered coding agents. 🧵
25
63
10
316
51,170
Hello, we are a research team at Northeastern University conducting a paid remote study on human–AI collaboration in coding workflows. We are currently looking for developers to participate in testing these workflows. Eligibility: Software engineering experience in software or website development Familiar with AI-assisted coding tools (e.g., GitHub Copilot, Cursor) Proficient with GitHub/GitLab, VS Code, and code review workflows Study Details: Collaborate with an AI agent to develop and review code (~4 hours in total) Compensation ($25/hr) via digital gift cards Participation is entirely voluntary. Interested? Sign up here: forms.gle/sUuFbyn5KnGZVd1N7 Contact me if you have any questions!
2
6
3
12
2,154
🎓 PhD Recruitment | Texas A&M CS I am seeking 1–2 PhD students for Fall 2026 in Human–AI Interaction, with a focus on AI for education. ✅ Motivated and responsible ✅ CS background ✅ Full-stack development or LLM fine-tuning experience is a plus 📅 Deadline: Dec 15, 2025
7
56
1
235
19,258
I rushed to move to the US on Saturday, so it is a pity that I cannot attend either UIST or Ubicomp this year. Hope to meet HCI people in some future conferences! Feel free to help me spread the word that I am looking for new PhD students to start my lab at Utah! #UIST #Ubicomp
1
4
53
5,587
#CHI2025 We will present "From Following to Understanding: Investigating the Role of Reflective Prompts in AR-Guided Tasks to Promote User Understanding" led by @nandizhang_ and @YukangYan PDF: dl.acm.org/doi/10.1145/37065… Video: youtu.be/QH7lpkRQcs0 Talk: 4/29 (Tue) 11:10-12:40
5
2
54
7,310
If you’re thinking about doing a PhD at the intersection of #design and #technology, this is the time — I’m hiring! In the Interactive 3D Design Lab, we’ll explore emerging interactive 3D design technologies such as real-time 3D, generative 3D, and spatial computing.
3
14
1
67
9,958
The sense of touch is innately private, but what if it could be shared across individuals? We @StanfordVR studied shared body sensations in VR and found it increases body illusion and empathy toward others and influences social behavior. #ISMAR2024 🔗vhil.stanford.edu/publicatio…
19
1
100
13,038
Thank you to everyone at #UIST2024 who attended my talk on my DrawTalking project or otherwise supported it. You’re the best!: xitter.cf/karltrosenberg/status/… Check back on my webpage for updates ktrosenberg.github.io/drawta… I’ll release the code and will post when that gets set-up.
Karl Toby Rosenberg@KarlTRosenberg9 Oct 2024
Introducing my #UIST2024 paper & demo: #DrawTalking: Building Interactive Worlds by Sketching and Speaking /w @AdobeResearch / @Adobe : @rubaiat @liyiwei , @HaijunXia Paper & Demos: ktrosenberg.github.io/drawta… Program: programs.sigchi.org/uist/202…
1
1
16
3,973
We have two papers that will be presented at VIS 2024! 1. StuGPTViz: A Visual Analytics Approach to Understand Student-ChatGPT Interactions 2. VisTellAR: Embedding Data Visualization to Short-form Videos Using Mobile Augmented Reality (YouTube link: youtu.be/JA0EX1fbyP0)
3
13
1,111
What if you could make physics diagrams come alive? At #UIST2024, we will be presenting our paper, Augmented Physics, an ML-Integrated Authoring Tool for Creating Interactive Physics Simulations from Static Diagrams Co-authors: @Freya_Wyy @nandizhang_ Jarin @rubaiat @ryosuzk
104
1,038
155
5,558
414,218
#UIST2023 We present HoloBots = Mobile Robots 🤖+ Holoportation 🥽 for physical telepresence in #MR (using #toio + #Hololens2 + #Kinect). This is led by @KeiichiIhara @MehradFaridan at @UCalgary and @UNIV_TSUKUBA_JP Video: youtu.be/KSBPtiXy8Hg Paper: arxiv.org/pdf/2307.16114.pdf
16
4
109
15,447
Load more