Joseph Chee Chang
@josephcc
😉💻🔄 Research Scientist @ AI2/Semantic Scholar | Artisanal small-batch handcrafted tweets with no added llms. @[email protected] @josephc.bsky.social
คุณอาจชื่นชอบ
Introducing Asta—our bold initiative to accelerate science with trustworthy, capable agents, benchmarks, & developer resources that bring clarity to the landscape of scientific AI + agents. 🧵
Producing reasoning texts boosts the capabilities of AI models, but do we humans correctly understand these texts? Our latest research suggests that we do not. This highlights a new angle on the "Are they transparent?" debate: they might be, but we misinterpret them. 🧵
People at #ACL2025, come drop by our poster today & chat with me about how context matters for reliable language model evaluations! Jul 30, 11:00-12:30 at Hall 4X, board 424.
Excited to share ✨ Contextualized Evaluations ✨! Benchmarks like Chatbot Arena contain underspecified queries, which can lead to arbitrary eval judgments. What happens if we provide evaluators with context (e.g who's the user, what's their intent) when judging LM outputs? 🧵↓
Ai2 is excited to be at #ACL2025 in Vienna, Austria this week. Come say hello, meet the team, and chat about the future of NLP. See you there! 🤝📚
Thank you all for joining us!! The Q&A were all very insightful 😀😀 Here's the link to the slides: bit.ly/acl25-hai-team
In Vienna for #ACL2025NLP this week! @josephcc, @aps6992 and I will present the Ai2 ScholarQA scientific QA system on Wed. I’ll also be at @sdpworkshop on Thurs! Hit me up if you’d like to chat about agents for science and post-training, or explore cafes in Vienna 🥐
We all agree that AI models/agents should augment humans instead of replace us in many cases. But how do we pick when to have AI collaborators, and how do we build them? Come check out our #ACL2025NLP tutorial on Human-AI Collaboration w/ @Diyi_Yang @josephcc, 📍7/27 9am@ Hall N!
In our new paper, “Contextualized Evaluations: Judging Language Model Responses to Underspecified Queries,” we find that adding just a bit of missing context can reorder model leaderboards—and surface hidden biases. 🧵👇
This new ScholarQA capability works for most openly licensed papers. It’s part of our commitment to transparency in science and making it easier to verify, trace, and build trusted AI.
You can now jump from Scholar QA answers to highlighted evidence in the source paper's pdf : )
Introducing SciArena, a platform for benchmarking models across scientific literature tasks. Inspired by Chatbot Arena, SciArena applies a crowdsourced LLM evaluation approach to the scientific domain. 🧵
@allen_ai @SemanticScholar is hiring an #ml #nlp #ai reasoning researcher for a Research Scientist, Agents for Science position with target start dates in 2025. Excited about developing AI systems with deep reasoning capabilities for science? Send an application our way!
🆕paper: LLMs Get Lost in Multi-Turn Conversation In real life, people don’t speak in perfect prompts. So we simulate multi-turn conversations — less lab-like, more like real use. We find that LLMs get lost in conversation. 👀What does that mean? 🧵1/N 📄arxiv.org/abs/2505.06120
I am presenting IdeaSynth in the last paper session at #CHI2025 right now! Feel free to come by G314-315 to learn about how we utilize LLM to provide literature-grounded assistance for research idea development! The talk is happening at around 10:12 AM.
🔬Research ideation is hard: After the spark of a brilliant initial idea, much work is still needed to further develop it into a well-thoughtout project by iteratively expanding and refining the initial idea and grounding it to relevant literature. How can we better support this?
AI agents are entering online social spaces, but often their messages feel generic or intrusive. In our #CHI25 paper, we introduce Social-RAG, a workflow that grounds AI generations in the specific group context by retrieving from the group’s interaction history. 🧵(1/9)
Meet Ai2 Paper Finder, an LLM-powered literature search system. Searching for relevant work is a multi-step process that requires iteration. Paper Finder mimics this workflow — and helps researchers find more papers than ever 🔍
We are looking for CS researchers to participate in a study exploring how AI can change the way we do literature reviews. 📚🧑🎓 Time: ~90 min, remote Compensation: $60 USD Sign up here: forms.gle/Pzw6YUhVUaZsS6… @dsweld @amyxzh @josephcc @marissa_rad @Siangliulue @turingmusician
We’re excited to share some updates to Ai2 ScholarQA: 🗂️ You can now sign in via Google to save your query history across devices and browsers. 📚 We added 108M+ paper abstracts to our corpus - expect to get even better responses! ✨ The backbone model has been updated to the…
working with PDFs for LMs, errors in extracting content can be soul-crushing we need not just clean OCR but also correct reading order and structured info reprsentation our latest in tools for this, olmOCR is incredible! here's page 5 of our tech report as plain text w/ table🤯
We took our most efficient model and made an open-source iOS app📱but why? As phones get faster, more AI will happen on device. With OLMoE, researchers, developers, and users can get a feel for this future: fully private LLMs, available anytime. Learn more from @soldni👇
United States เทรนด์
- 1. Columbus 158K posts
- 2. President Trump 1.1M posts
- 3. Middle East 259K posts
- 4. Brian Callahan 9,354 posts
- 5. #IndigenousPeoplesDay 11.3K posts
- 6. Titans 40.8K posts
- 7. Thanksgiving 56K posts
- 8. Vrabel 7,403 posts
- 9. Cape Verde 11.4K posts
- 10. Seth 49.5K posts
- 11. Macron 219K posts
- 12. Marc 50.4K posts
- 13. #WWERaw 55.1K posts
- 14. HAZBINTOOZ 5,085 posts
- 15. Shildt 2,804 posts
- 16. Apple TV 5,522 posts
- 17. Sarah Strong 1,891 posts
- 18. Egypt 268K posts
- 19. #Isles 1,328 posts
- 20. Darius Smith 3,865 posts
คุณอาจชื่นชอบ
-
Yoonjoo Lee @ COLM 2025
@yoonjoo_le2 -
Michelle Lam
@michelle123lam -
Michael Xieyang Liu
@lxieyang -
Qian Yang
@fabulousQian -
Juho Kim
@imjuhokim -
Pao Siangliulue
@Siangliulue -
Jonathan Bragg
@turingmusician -
yuwen lu
@yuwen_lu_ -
Toby J. Li😺 (he/him)
@TobyJLi -
Bryan Wang
@bryanhaoenwang -
Vera Liao
@QVeraLiao -
Mitchell Gordon
@mitchellgordon -
Kenneth Huang
@windx0303 -
Anhong Guo
@AnhongGuo -
Gagan Bansal
@bansalg_
Something went wrong.
Something went wrong.