At Athens this week to present my work at #FAccT2025! 🏛️ I'll be showcasing the work on understanding the ethics behind generative search engines or deep research models. If you're around or would like to grab coffee, feel free to DM me! ☺️☕ Arxiv link: arxiv.org/abs/2410.22349

That random moment at a conference when you realise your badge doesn't say "PhD student" anymore. Just your name and your new affiliation. Feels good, feels weird. 😊😊 Seeing the academic world from this new side hits different for sure. #Facct2025

🔍 Outstanding engagement at our Search Engines in the AI Era presentation! 🎉 Great discussions with the #ACMFAccT2025 community about the critical gap between promised factual accuracy and actual performance in LLM-based search systems. When AI-powered search becomes

Thank you ACM FAccT! Presenting my work here was extremely fun. Such a great place for people who conduct interdisciplinary research in AI, ethics and more. 😊🏛️ First venue where I saw researchers, policy makers and industry folks coming together for a conversation.

Didn't realize that my citations had gone above 500! Oh wow! 🎆🙀 Kinda speechless by the attention my works have gotten recently. This speaks more of how ethical and sociotechnical NLP are gaining importance! So grateful for everyone in this field! Thank you! 😊 #NLProc #AI

🌟 Excited to present our work at Empirical Methods in Natural Language Processing EMNLP 2025 - a leading conference in NLP and AI research! 📄 Our accepted papers: Topic-Guided Reinforcement Learning with LLMs for Enhancing Multi-Document Summarization 👥Authors: Chuyuan Li

EACL 2026 website is live and Call for Papers is out! Join us at #EACL2026 (Rabat, Morocco 🇲🇦, Mar 24-29 2026) 👉 Open to all areas of CL/NLP + related fields. Details: 2026.eacl.org/calls/papers/ • ARR submission deadline: Oct 6, 2025 • EACL commitment deadline: Dec 14, 2025

EMNLP 2025 / #EMNLP2025 Accepted Paper: Direct Judgement Preference Optimization 📝 Paper: arxiv.org/abs/2409.14664 This work introduces SFR-Judges, a family of generative judge models trained with Direct Preference Optimization (DPO) to enhance LLM evaluation capabilities

So so grateful that my research was featured in Penn State Research 2024 + IST’s iConnect 2025, where I finished my PhD! 🙌🥳 My work blurb: studying the sociotechnical impact of generative AI Now at Salesforce AI Research , I continue that journey 🚀 #AI #NLProc #TrustworthyAI #Research

I am looking for students and collaborators for several new projects at the intersection of AI, the economy, and society. If you are interested, please fill out this form! forms.gle/DNrKPeiJjdi4Vg…

AI-powered deep research agents promise trustworthy, source-grounded research. But do they actually deliver? 🧐 Our new paper introduces DeepTRACE, the first sociotechnical audit to measure how reliable these systems really are! Salesforce AI Research Paper: arxiv.org/pdf/2509.04499 1/n

Meet SFR-DeepResearch (SFR-DR) 🤖: our RL-trained autonomous agents that can reason, search, and code their way through deep research tasks. 🚀SFR-DR-20B achieves 28.7% on Humanity's Last Exam (text-only) using only web search 🔍, browsing 🌐, and Python interpreter 🐍,

As deep research AI become more widely used, it’s important to evaluate their impact on society! I'm grateful to see our recent work that looks at this is receiving the attention it deserves! 🎉📰 New Scientist: newscientist.com/article/249613… Tech Xplore: techxplore.com/news/2025-09-a…

🚨BREAKING: MIT researchers discover how to enable LLMs to do real logical reasoning. This is what you need to know: (thread)

Introducing Claude Sonnet 4.5—the best coding model in the world. It's the strongest model for building complex agents. It's the best model at using computers. And it shows substantial gains on tests of reasoning and math.