Combining Twitter/X Scientific Discussion With Paper Reading Experience
Papers often start discussions on social media (especially Twitter/X and Bluesky) between authors, and field experts. These conversations often clarify hand-wavy points in the paper and reveal its limitations more directly than the paper itself does.
My colleagues and I built a research prototype that integrates these conversations into the paper reading experience. It retrieves relevant social media discussions about a paper and presents them alongside it, with double-sided synchronization so you can see which parts of the paper a discussion relates to and which discussions exist for any given section. We published this work at UIST 2025.
We've already added 8 papers from ICML, ICLR, NeurIPS, and COLM as a showcase. The screenshot is for "Position: AI/ML Influencers Have a Place in the Academic Process" (ICML'24). We'd love to hear your thoughts and feedback.
Check it out here: https://aceatusc.github.io/surf/