Hanbo Xie

1K posts

Hanbo Xie

Hanbo Xie

@PsychBoyH

Ph.D. student in @NRDLab in @GeorgiaTech. Alum in @uarizona @PKU1898 @cocosci_lab. Co-Founder of @Mind_RL_Hub. Human and AI thinking and reasoning.

Atlanta Katılım Mayıs 2020
1.7K Takip Edilen864 Takipçiler
Sabitlenmiş Tweet
Hanbo Xie
Hanbo Xie@PsychBoyH·
I wrote a blog post about my recent thoughts on the scaling in Computational Cognitive Sciences and AI (CogAI). In this blog, I argue that performance scales more easily than insights by pointing out two bottlenecks. Thoughts and comments are welcome! xhb120633.github.io/blog/performan…
Hanbo Xie tweet media
English
1
10
29
2K
Hanbo Xie retweetledi
Anna Ivanova
Anna Ivanova@neuranna·
Language, Intelligence & Thought lab is looking for a lab manager! This is a 2-year postbac position that will allow you to gain experience in human neuroscience, cognitive science, and AI research prior to applying to PhD programs. Express interest here: forms.gle/289sLgZdJ2bQr1…
English
1
27
55
8.4K
kkpk
kkpk@iv76954102·
@PsychBoyH Well, Centaur is a joke, a lot has already been written about it.
English
1
0
0
31
Hanbo Xie
Hanbo Xie@PsychBoyH·
I wrote a blog post about my recent thoughts on the scaling in Computational Cognitive Sciences and AI (CogAI). In this blog, I argue that performance scales more easily than insights by pointing out two bottlenecks. Thoughts and comments are welcome! xhb120633.github.io/blog/performan…
Hanbo Xie tweet media
English
1
10
29
2K
Hanbo Xie
Hanbo Xie@PsychBoyH·
@iv76954102 As a blog, I tried to keep it lightweight so I left out many examples from my original draft. But here is probably the most well-known one in the last year about predicting behavior. nature.com/articles/s4158…
English
1
0
0
33
kkpk
kkpk@iv76954102·
@PsychBoyH I don't really see any progress in predicting behaviour in the last years. What I see is a growing number of stupid correlational studies which are practically and scientifically useless. It would be good to include some examples in your post, right now it is rather vague.
English
1
0
0
37
Ionatan Kuperwajs
Ionatan Kuperwajs@Ikuperwajs·
New little family member 🐈‍⬛🖤
Ionatan Kuperwajs tweet media
English
1
0
5
136
Hanbo Xie
Hanbo Xie@PsychBoyH·
@jlambert_ Hi John, I just came across this post and submitted the express of interests. I have completed research interview and am in the phase of team matching. If the hiring process is still active, I would be appreciated to be considered!
English
0
0
0
30
John Lambert
John Lambert@jlambert_·
I am hiring a Student researcher at Google DeepMind for Winter/Spring or Summer of 2026! 🛠️ Interested in tool-use and reasoning for LLMs? 🚀 Up-to-date with frontier research? 🔬 Love to understand code & algorithms from first principles? 📚 Currently studying for a BS/MS/PhD?
John Lambert tweet media
English
34
86
1.2K
90.1K
Hanbo Xie retweetledi
Griffiths Computational Cognitive Science Lab
Griffiths Computational Cognitive Science Lab@cocosci_lab·
Excited to announce a new book telling the story of mathematical approaches to studying the mind, from the origins of cognitive science to modern AI! The Laws of Thought will be published in February, and is available for pre-order now.
Griffiths Computational Cognitive Science Lab tweet media
English
36
262
1.7K
94.6K
Hanbo Xie
Hanbo Xie@PsychBoyH·
We have a poster @NeurIPSConf right this afternoon! Location: 2200, Exhibit Hall C, D & E Time: 16:00 - 19:00 PST Summary: We use cog sci insights to find failures of LLMs in open-ended exploration due to insufficient reasoning. You are welcome to stop by our poster and connect.
Hanbo Xie tweet media
English
0
0
6
213
Hanbo Xie
Hanbo Xie@PsychBoyH·
@AndrewLampinen Nice to see that! Andrew! You are welcome to our poster on Thursday about LLM exploration and reasoning. I would like to have a 15-min chat if you have available slots. Happy to reconnect at NeruIPS soon!
English
0
0
2
156
Andrew Lampinen
Andrew Lampinen@AndrewLampinen·
Heading to NeurIPS this week! Let me know if you want to chat about science of what models learn, interpretability, what models learn in context vs. from their training data, etc. A few things I'm involved in:
English
25
4
163
14.4K
Hanbo Xie
Hanbo Xie@PsychBoyH·
I am excited to share that our work is accepted by @NeurIPSConf . This is an interesting work using insights and tools from comp cog neuroscience to understand LLMs. Nice work co-led by @LanPan1208 and @NRDlab !
Hanbo Xie@PsychBoyH

Large Language Models can do a lot of things. But do you know they cannot explore effectively, especially in open-ended tasks? Recently, @LanPan1208 @NRDlab and I dropped a preprint to investigate how LLMs explore in an open-ended task. arxiv.org/abs/2501.18009

English
1
1
12
1.3K
Hanbo Xie
Hanbo Xie@PsychBoyH·
This paper cannot be done without my supreme collaborator and supervisor, @HuaDongXiong and @NRDlab. I also want to acknowledge the valuable discussions with @JQ_Zhu and @JingmingXue during the draft phase, and our excellent RA teams who are not on academia Twitter.
English
1
0
4
305
Hanbo Xie
Hanbo Xie@PsychBoyH·
Can Think-Aloud be really useful in understanding human minds? Building on our previous work, we formally propose reopening this old debate, with one of the largest Think-Aloud datasets, "RiskyThought44K," and LLM analysis, showing Think-Aloud can complement to comp cogsci.
Hanbo Xie tweet media
English
1
3
33
2.5K

Keşfet

@iv76954102 @Ikuperwajs @iclr_conf @jlambert_ @_rdgao @NeurIPSConf @AndrewLampinen @StePalminteri