A CV/ML researcher. Assoc. Prof. @SeoulNatlUni. Prev: @allen_ai @MSFTResearch @Adobe and @UofMaryland Ph.D. grad. He/him

Seoul, South Korea
Joined May 2009
Jonghyun Choi retweeted
Check out my first music video 🏖️ Plus, accepting applications for hot (robo)secretaries Want yours?
99
51
39
338
142,689
Jonghyun Choi retweeted
Congratulations to everyone whose papers were accepted to ECCV 2026! 🎉 We are organing the Embodied Agents and Dialog (EAD) Workshop at ECCV 2026. We look forward to your participation! More details: ead-workshop.github.io/ #ECCV2026 #EmbodiedAI #Dialog
1
1
2
870
Jonghyun Choi retweeted
We are so back!
People talk, listen, watch, think, and collaborate at the same time, in real time. We've designed an AI that works with people the same way. We share our approach, early results, and a quick look at our model in action. thinkingmachines.ai/blog/int…
36
17
4
545
55,467
Jonghyun Choi retweeted
Our position paper “Modular Memory is the Key to Continual Learning Agents” (arxiv.org/abs/2603.01761) has been accepted to #ICML2026 @icmlconf as a spotlight! 🎊 Read for a modern perspective on memory, continual learning, and sustainable adaptation at foundation model scale!
Following our @dagstuhl seminar on Continual Learning in the Foundation Model Era, we are now sharing a roadmap! tldr: we view modular memory design as the missing piece to combine the capabilities of In-Weight & In-Context Learning for adaptation at scale arxiv.org/abs/2603.01761
5
29
166
15,323
Jonghyun Choi retweeted
After many fantastic years at Ai2, I am so excited to say I have started a new company with some of the most intelligent, driven, and joyful people I know: Kiana (@ehsanik), Ross (@inkynumbers), and Matt (@mattdeitke).
🚨BREAKING: A mysterious virus is spreading. Unfortunately, Ross Girshick (@inkynumbers), @mattdeitke, @LucaWeihs, and I recently got infected. 🦠 This brain-targeting virus causes highly irrational behavior. Common symptoms include: 🛑 Quitting well-paying jobs 🛑 Leaving behind academic awards 🛑 Diving headfirst into the chaos of uncertainty There’s no cure in sight! Side effects: A dangerous amount of FUN. Stay safe, folks… or don’t. You might just end up building something amazing! 🚀 Do you have a new year’s resolution to quit your job? We are hiring, [email protected]! Wanna know what we are up to? Sign up for updates here: vercept.com
19
6
87
13,177
Jonghyun Choi retweeted
📢 Don't miss the 5th workshop on Real-World Surveillance: Applications & Challenges at #WACV2025 in Arizona. 🏜️ Submit your papers by Nov 27th, 2024. More details here:
We are very happy to run the 5th version of our workshop on Real-World Surveillance: Applications and Challenges at #WACV2025, in Arizona. Updated Paper submission deadline: Nov. 27th, 2024. More details can be found here: vap.aau.dk/rws-wacv2025/ @wacv_official
1
4
1,774
Jonghyun Choi retweeted
📢 Exciting News! The Fourth Conference on Lifelong Learning Agents (CoLLAs 2025) will be held at the University of Pennsylvania (@Penn) in Philadelphia, USA 🇺🇸 🗓️ Important Dates: Abstract Deadline: Feb 21, 2025 Submission Deadline: Feb 26, 2025 Conference Dates: Aug 11 - Aug 14, 2025 We invite submissions that present new theories, methodologies, applications, or insights into algorithms and benchmarks designed for non-i.i.d. and non-stationary settings. Accepted papers will be published in the Proceedings of Machine Learning Research (PMLR). 📚 Full CFP: lifelong-ml.cc/Conferences/2… #CoLLAs2025 #AI #MachineLearning #ContinualLearning #LifelongLearning #ResearchConference #CallForPapers #NonStationaryLearning
1
25
14
70
57,687
Jonghyun Choi retweeted
CoRL 2025 will be co-located with Humanoids 2025 at the same venue!
1
1
18
9,004
Jonghyun Choi retweeted
I am extremely pleased to announce that CoRL 2025 will be in Seoul, Korea! The organizing team includes myself and @gupta_abhinav_ as general chairs, and @JosephLim_AI, @songshuran, and Hae-Won Park (KAIST) as program chairs.
4
14
2
170
26,467
We are thrilled to announce that the next CoRL conference will happen in Seoul! #CoRL #Robotics #learning #AI
I am extremely pleased to announce that CoRL 2025 will be in Seoul, Korea! The organizing team includes myself and @gupta_abhinav_ as general chairs, and @JosephLim_AI, @songshuran, and Hae-Won Park (KAIST) as program chairs.
15
98
10,687
Jonghyun Choi retweeted
🚨ReALFRED: An Embodied Instruction Following Benchmark in Photo-Realistic Environments [ECCV'24] 🌟𝐏𝐫𝐨𝐣: twoongg.github.io/projects/r… 🚀𝐀𝐛𝐬: arxiv.org/abs/2407.18550 propose the ReALFRED benchmark that employs real-world scenes, objects, and room layouts to learn agents to complete household tasks by understanding free-form language instructions and interacting with objects in large, multi-room and 3D-captured scenes.
3
8
535
Jonghyun Choi retweeted
📢 The wait is over – paper decisions have been sent out! 🎉 Check your inbox to see if your hard work paid off. Congrats to everyone, whether your paper was accepted or not. This is just one step in the journey – keep pushing the boundaries of knowledge! 🚀 #Research
3
12
2
111
21,416
Jonghyun Choi retweeted
🎉 We just sent out notifications for the first round of #WACV2025! Here are some interesting stats: 📄 1381 submissions ❌ 504 papers were rejected ~36.5% 🔄 710 papers were invited for revision ~51.4% ✅ 167 papers were accepted ~12.1% On to the next round! 🚀
2
7
1
59
9,918
We already have multiple small conferences that have much more high quality review process like @RL_Conference, @CoLLAs_Conf, @automl_conf, @LogConference, @Conf_CLeaR. I think people should start submitting to these high quality venues than trying the lottery at NeurIPS.
IMO, it's time to break down NeurIPS/ICML into smaller tracks to improve this and also the reviewer-paper matching. Authors would pick 1-2 tracks at submission, and reviewers would be recruited by tracks. Essentially we would have multiple smaller conferences running in parallel.
3
6
35
5,109
So proud of the entire OLMo team. OLMo 1.0 was such a nice team effort, advocating for fully open LMs. Better models and papers coming soon!
🏆 ACL Best Theme Paper Award: - OLMo: Accelerating the Science of Language Models by Groeneveld et al. #NLProc #ACL2024NLP
1
1
46
2,957
Jonghyun Choi retweeted
🏆 ACL Best Theme Paper Award: - OLMo: Accelerating the Science of Language Models by Groeneveld et al. #NLProc #ACL2024NLP
10
1
112
15,613
Jonghyun Choi retweeted
#CoLLAs2024 was such a small conference, yet I had so many great discussions these past days! And I met such a very friendly community 🐨 Kudos to the organizers !! 🎉
#CoLLas2024 folks 🐨 Temperature rising for the coming sessions 🔥🐨
2
20
1,169
Jonghyun Choi retweeted
📢3 DAYS LEFT UNTIL #CoLLAs2024 🥳 We cannot wait for you to join us next week in Pisa at the International Conference on Lifelong Learning Agents! See you soon! 👋 Stay up to date with the schedule 👇 lifelong-ml.cc/Conferences/2…
3
15
740
Language is never spoken in isolation. It is expressed and contextualized within the speaker’s actions and environment. We introduce an embodied instruction following dataset addressing the inherent underspecification and ambiguity of communication and evolving intents of humans.
We introduce Situated Instruction Following (SIF), to appear in ECCV 2024! There is inherent underspecification in instructions when humans act as they speak. SIF is addresses these dynamic, temporally evolving intent of instructions in the context of physical human actions.(1/7)
2
5
921
Jonghyun Choi retweeted
Thrilled to announce latest work in the VisProg research direction at @allen_ai with amazing co-authors @LucaWeihs @anikembhavi CodeNav: Beyond tool-use to using real-world codebases with LLM agents TL;DR Code-use > Tool-use: Given a codebase it has never seen before, our CodeNav agent searches functions/classes in the codebase to solve a complex user query. Our agent not only generates free-form code, but executes it, gets feedback, and corrects errors. ❌Tool-description ❌Tool-demonstrations ✅Zero-shot code-use with a high-level library description (like a README) Project Page: codenav.allenai.org/ Paper: arxiv.org/abs/2406.12276v1 Article: 👇
3
2
35
6,787