Danqing Shi @danqingshi.bsky.social · 06/02/2026Better AI models by incorporating human feedback into training #RLHF fcai.fi/news/2026/2/...fcai.fiBetter AI models by incorporating user feedback into training — FCAINew research improves a popular method for fine-tuning AI models by 60% using visualization tools. 000
Danqing Shi @danqingshi.bsky.social · 01/12/2025Feedback in RLHF is collected from human raters who are prompted by the system to compare pairs and express preferences. What advantages do we get if humans can exhibit higher agency? doi.org/10.1111/cgf.... 121
Danqing Shi @danqingshi.bsky.social · 17/09/2025Thrilled to share our #UIST2025 research! We investigate how the decomposition principle can improve human feedback for LLM alignment. In a 160-participant study, our tool DxHF increases feedback accuracy by +4.7% 👉 sdq.github.io/DxHF Furui Tino @oulasvirta.bsky.social @elassady.bsky.social 021
Reposted by Danqing ShiAntti Oulasvirta @oulasvirta.bsky.social · 20/08/2025📢The open access version of our book is available now via OUP's site: global.oup.com/academic/pro... 0279
Danqing Shi @danqingshi.bsky.social · 14/05/2025AI models are paving the way for explainable AI and better human-computer interaction fcai.fi/news/2025/4/...fcai.fiAI that mirrors how humans behave can drive better designs for keyboards and charts — FCAIThe human-like performance of these AI models is also transparent, paving the way for explainable AI and better human-computer interaction. 010
Danqing Shi @danqingshi.bsky.social · 27/03/2025Our paper has been selected for #CHI2025 Best Paper Honorable Mention recognition 🥳🥳 020
Danqing Shi @danqingshi.bsky.social · 27/02/20251/ Why do people make so many errors in touchscreen typing, and how do people fix them? Our #CHI2025 paper introduces Typoist, the computational model to simulate human errors across perception, motor, and memory. 📄 arxiv.org/abs/2502.03560 131
Reposted by Danqing ShiAntti Oulasvirta @oulasvirta.bsky.social · 26/02/2025Many thanks to SIGCHI for recognizing our work and to numerous brilliant colleagues. It is a great honor to join the Academy. 3172
Danqing Shi @danqingshi.bsky.social · 24/02/20251/ How people read charts when they have a specific task in mind? Their gaze isn’t random! Our #CHI2025 paper introduces Chartist, the first model designed to simulate these task-driven eye movements. 📄 arxiv.org/abs/2502.03575 141
