Charley Wu | hiring PhDs/Postdocs
@thecharleywu.bsky.social
Prof. of Computational Cognitive Science at TU Darmstadt & PI of the Human and Machine Cognition lab at the University of Tübingen | hmc-lab.com
Just 1 week to apply! 4 year @erc.europa.eu funded PhD position working in an interdisciplinary team to study #culturalEvolution as a process of reuse, recombination, and creative re-engineering of past solutions. Details 👉 hmc-lab.com/ERCPhDCultur... 🙏Please share!
November 5, 2025 at 10:24 AM
Just 1 week to apply! 4 year @erc.europa.eu funded PhD position working in an interdisciplinary team to study #culturalEvolution as a process of reuse, recombination, and creative re-engineering of past solutions. Details 👉 hmc-lab.com/ERCPhDCultur... 🙏Please share!
Great to see this work finally released! Fun fact, Valerii was the winner of the #COSMOS2023 poster prize with an earlier iteration of this project
October 15, 2025 at 3:42 PM
Great to see this work finally released! Fun fact, Valerii was the winner of the #COSMOS2023 poster prize with an earlier iteration of this project
Me and @watarutoyokawa.bsky.social want to thank everyone who attended, presented at, and helped organize #COSMOS2025! If you weren't able to make it, all teaching materials are already online and we'll be uploading videos of all talks soon cosmossummerschool.github.io/materials/
October 6, 2025 at 4:15 AM
Me and @watarutoyokawa.bsky.social want to thank everyone who attended, presented at, and helped organize #COSMOS2025! If you weren't able to make it, all teaching materials are already online and we'll be uploading videos of all talks soon cosmossummerschool.github.io/materials/
Mentorship sessions are one of my favorite parts of #COSMOS2025 Discussing science and academic careers, while exploring what Tokyo has to offer! With @liang-lee.bsky.social @mohsen-raoufi.bsky.social @wasita.bsky.social @shannonyasuda.bsky.social @kefang.bsky.social @davidschultner.bsky.social
October 2, 2025 at 4:49 AM
Mentorship sessions are one of my favorite parts of #COSMOS2025 Discussing science and academic careers, while exploring what Tokyo has to offer! With @liang-lee.bsky.social @mohsen-raoufi.bsky.social @wasita.bsky.social @shannonyasuda.bsky.social @kefang.bsky.social @davidschultner.bsky.social
As our last invited speaker of #COSMOS2025, @lazytenuredprof.bsky.social uses sequential photos of Andrey Markov as an introduction to how MCMC with people can be used to understand (cultural) knowledge
October 2, 2025 at 12:48 AM
As our last invited speaker of #COSMOS2025, @lazytenuredprof.bsky.social uses sequential photos of Andrey Markov as an introduction to how MCMC with people can be used to understand (cultural) knowledge
Using social reinforcement learning to figure out which japanese konbini (convenience store) is superior. Where are you getting your strong zero, onigiri, or egg sandwich?
Last #COSMOS2025 talk of the day by Shinsuke Suzuki sites.google.com/view/szklab/...
Last #COSMOS2025 talk of the day by Shinsuke Suzuki sites.google.com/view/szklab/...
October 1, 2025 at 7:41 AM
Using social reinforcement learning to figure out which japanese konbini (convenience store) is superior. Where are you getting your strong zero, onigiri, or egg sandwich?
Last #COSMOS2025 talk of the day by Shinsuke Suzuki sites.google.com/view/szklab/...
Last #COSMOS2025 talk of the day by Shinsuke Suzuki sites.google.com/view/szklab/...
A small selection of #COSMOS2025 participants hard at work on their group projects @lazytenuredprof.bsky.social @damienfarine.bsky.social @lucymaplin.bsky.social Thanks to all of the instructors for providing code/data for hands-on experience with modeling cosmossummerschool.github.io/projects/
October 1, 2025 at 5:51 AM
A small selection of #COSMOS2025 participants hard at work on their group projects @lazytenuredprof.bsky.social @damienfarine.bsky.social @lucymaplin.bsky.social Thanks to all of the instructors for providing code/data for hands-on experience with modeling cosmossummerschool.github.io/projects/
Now up at #COSMOS2025: @kartikchandra.bsky.social & Lio Wong giving a tutorial on recursive social reasoning using MEMO github.com/kach/memo
Fun fact: those flowers between them and 🗻 are called "cosmos".
Collab notebook here to follow along 👉 cosmossummerschool.github.io/materials/#g...
Fun fact: those flowers between them and 🗻 are called "cosmos".
Collab notebook here to follow along 👉 cosmossummerschool.github.io/materials/#g...
October 1, 2025 at 2:05 AM
Now up at #COSMOS2025: @kartikchandra.bsky.social & Lio Wong giving a tutorial on recursive social reasoning using MEMO github.com/kach/memo
Fun fact: those flowers between them and 🗻 are called "cosmos".
Collab notebook here to follow along 👉 cosmossummerschool.github.io/materials/#g...
Fun fact: those flowers between them and 🗻 are called "cosmos".
Collab notebook here to follow along 👉 cosmossummerschool.github.io/materials/#g...
Shifting the spotlight from dominance towards shared decision-making in collective groups @damienfarine.bsky.social #COSMOS2025
September 30, 2025 at 7:38 AM
Shifting the spotlight from dominance towards shared decision-making in collective groups @damienfarine.bsky.social #COSMOS2025
Up now at #COSMOS2025 @lucymaplin.bsky.social presenting work on social learning and cultural inheritance in wild great tits 🐦
September 30, 2025 at 4:44 AM
Up now at #COSMOS2025 @lucymaplin.bsky.social presenting work on social learning and cultural inheritance in wild great tits 🐦
Key dimensions of social learning problems. @watarutoyokawa.bsky.social giving the first tutorial of #COSMOS2025 As always, course materials are opening available on our website cosmossummerschool.github.io/materials/
September 30, 2025 at 2:19 AM
Key dimensions of social learning problems. @watarutoyokawa.bsky.social giving the first tutorial of #COSMOS2025 As always, course materials are opening available on our website cosmossummerschool.github.io/materials/
Tatsuya Kameda giving the opening keynote at #COSMOS2025 (photo care of @henrivdd.bsky.social)
September 30, 2025 at 1:27 AM
Tatsuya Kameda giving the opening keynote at #COSMOS2025 (photo care of @henrivdd.bsky.social)
🚀👩🚀 COSMOS 2025 is coming to Tokyo in less than a week! Me + @watarutoyokawa.bsky.social + our amazing faculty👇 can’t wait to welcome everyone to RIKEN.
📣 Attendees: check your inbox for info about travel, accommodations, mentorship groups, guest access, etc...
cosmossummerschool.github.io
📣 Attendees: check your inbox for info about travel, accommodations, mentorship groups, guest access, etc...
cosmossummerschool.github.io
September 23, 2025 at 9:51 AM
🚀👩🚀 COSMOS 2025 is coming to Tokyo in less than a week! Me + @watarutoyokawa.bsky.social + our amazing faculty👇 can’t wait to welcome everyone to RIKEN.
📣 Attendees: check your inbox for info about travel, accommodations, mentorship groups, guest access, etc...
cosmossummerschool.github.io
📣 Attendees: check your inbox for info about travel, accommodations, mentorship groups, guest access, etc...
cosmossummerschool.github.io
In Hegel's childhood home, they have an escape room where you have to steal the Phenomenology of Spirit by lamplight. It's for a very niche audience, but 10/10
September 20, 2025 at 2:37 PM
In Hegel's childhood home, they have an escape room where you have to steal the Phenomenology of Spirit by lamplight. It's for a very niche audience, but 10/10
ARR and RRR map onto different parameterizations of a Causal Bayes net:
ARR = linear + additive influence
RRR = noisy-AND-NOT (proportional risk reduction)
These yield divergent predictions when baseline risks vary (e.g. healthcare workers vs. general population).
ARR = linear + additive influence
RRR = noisy-AND-NOT (proportional risk reduction)
These yield divergent predictions when baseline risks vary (e.g. healthcare workers vs. general population).
September 4, 2025 at 9:08 AM
ARR and RRR map onto different parameterizations of a Causal Bayes net:
ARR = linear + additive influence
RRR = noisy-AND-NOT (proportional risk reduction)
These yield divergent predictions when baseline risks vary (e.g. healthcare workers vs. general population).
ARR = linear + additive influence
RRR = noisy-AND-NOT (proportional risk reduction)
These yield divergent predictions when baseline risks vary (e.g. healthcare workers vs. general population).
ARR and RRR are derived from the same data but they implement different ideas about causality.
We show ARR = ΔP, RRR = causal power. Both are measures of causal strength, but they rest on different assumptions.
We show ARR = ΔP, RRR = causal power. Both are measures of causal strength, but they rest on different assumptions.
September 4, 2025 at 9:08 AM
ARR and RRR are derived from the same data but they implement different ideas about causality.
We show ARR = ΔP, RRR = causal power. Both are measures of causal strength, but they rest on different assumptions.
We show ARR = ΔP, RRR = causal power. Both are measures of causal strength, but they rest on different assumptions.
🚨 Hot off the presses! 📢 During COVID, many asked: “How much safer am I after 1 vs. 2 vaccines?” Our new paper shows why the answer depends on how treatment effects are measured: absolute vs. relative risk reductions (ARR vs. RRR)
👉 doi.org/10.1016/j.jm...
led by bmeder.org & w/ Felix Rebitschek 👇🧵
👉 doi.org/10.1016/j.jm...
led by bmeder.org & w/ Felix Rebitschek 👇🧵
September 4, 2025 at 9:08 AM
🚨 Hot off the presses! 📢 During COVID, many asked: “How much safer am I after 1 vs. 2 vaccines?” Our new paper shows why the answer depends on how treatment effects are measured: absolute vs. relative risk reductions (ARR vs. RRR)
👉 doi.org/10.1016/j.jm...
led by bmeder.org & w/ Felix Rebitschek 👇🧵
👉 doi.org/10.1016/j.jm...
led by bmeder.org & w/ Felix Rebitschek 👇🧵
Proud of the work from hmc-lab.com & collaborators @ #CogSci2025 this year, but sad I cant be there myself
@hanqizhou.bsky.social @davidnagy.bsky.social @alexthewitty.bsky.social @stepalminteri.bsky.social @brendenlake.bsky.social @kefang.bsky.social @rdhawkins.bsky.social @meanwhileina.bsky.social
@hanqizhou.bsky.social @davidnagy.bsky.social @alexthewitty.bsky.social @stepalminteri.bsky.social @brendenlake.bsky.social @kefang.bsky.social @rdhawkins.bsky.social @meanwhileina.bsky.social
July 29, 2025 at 2:53 PM
Proud of the work from hmc-lab.com & collaborators @ #CogSci2025 this year, but sad I cant be there myself
@hanqizhou.bsky.social @davidnagy.bsky.social @alexthewitty.bsky.social @stepalminteri.bsky.social @brendenlake.bsky.social @kefang.bsky.social @rdhawkins.bsky.social @meanwhileina.bsky.social
@hanqizhou.bsky.social @davidnagy.bsky.social @alexthewitty.bsky.social @stepalminteri.bsky.social @brendenlake.bsky.social @kefang.bsky.social @rdhawkins.bsky.social @meanwhileina.bsky.social
Our pre #cogsci2025 workshop @unituebingen.bsky.social is wrapping up. Thx to @ml4science.bsky.social for supporting the event, @alexthewitty.bsky.social & Polina Tsvilodub for helping organize, and all the amazing participants who came from as far away as Tokyo!
July 24, 2025 at 1:07 PM
Our pre #cogsci2025 workshop @unituebingen.bsky.social is wrapping up. Thx to @ml4science.bsky.social for supporting the event, @alexthewitty.bsky.social & Polina Tsvilodub for helping organize, and all the amazing participants who came from as far away as Tokyo!
New poster for the local pre- #cogsci2025 workshop I am organizing at @unituebingen.bsky.social with generous support from @ml4science.bsky.social Please share widely! Accepted Cogsci papers are not required to present a talk/poster! Also feel free to just come and checkout some cool research!
June 18, 2025 at 2:48 PM
New poster for the local pre- #cogsci2025 workshop I am organizing at @unituebingen.bsky.social with generous support from @ml4science.bsky.social Please share widely! Accepted Cogsci papers are not required to present a talk/poster! Also feel free to just come and checkout some cool research!
Our 🇩🇪 lab is currently on a retreat in Strasbourg 🇫🇷 the capital of Europe 🇪🇺 Feat. workshops on computational psychiatry & program induction, plus many other great talks from lab members #ChooseEurope
June 12, 2025 at 3:30 PM
Our 🇩🇪 lab is currently on a retreat in Strasbourg 🇫🇷 the capital of Europe 🇪🇺 Feat. workshops on computational psychiatry & program induction, plus many other great talks from lab members #ChooseEurope
🚀Join our team @tuda.bsky.social ! 🚀
I'm looking for 3 PhDs & 1 Postdoc for my @erc.europa.eu project “C4: Compositional Compression in Cognition and Culture” to study learning across individuals, teams, and cultural timescales
👉 PhD: hmc-lab.com/ERC_PhDs.html
👉 Postdoc: hmc-lab.com/ERC_Postdoc....
I'm looking for 3 PhDs & 1 Postdoc for my @erc.europa.eu project “C4: Compositional Compression in Cognition and Culture” to study learning across individuals, teams, and cultural timescales
👉 PhD: hmc-lab.com/ERC_PhDs.html
👉 Postdoc: hmc-lab.com/ERC_Postdoc....
June 11, 2025 at 8:00 AM
🚀Join our team @tuda.bsky.social ! 🚀
I'm looking for 3 PhDs & 1 Postdoc for my @erc.europa.eu project “C4: Compositional Compression in Cognition and Culture” to study learning across individuals, teams, and cultural timescales
👉 PhD: hmc-lab.com/ERC_PhDs.html
👉 Postdoc: hmc-lab.com/ERC_Postdoc....
I'm looking for 3 PhDs & 1 Postdoc for my @erc.europa.eu project “C4: Compositional Compression in Cognition and Culture” to study learning across individuals, teams, and cultural timescales
👉 PhD: hmc-lab.com/ERC_PhDs.html
👉 Postdoc: hmc-lab.com/ERC_Postdoc....
Lastly, we propose variable-rate encoding as a new framework for memory: predictable info gets heavy compression (low fidelity), while surprising events are only minimally compressed (high fidelity). This trade-off ensures details survive to improve future models.
June 6, 2025 at 8:11 AM
Lastly, we propose variable-rate encoding as a new framework for memory: predictable info gets heavy compression (low fidelity), while surprising events are only minimally compressed (high fidelity). This trade-off ensures details survive to improve future models.
Our framework explains a host of classic phenomena: DRM false memories (schema-consistent lures slip in when compressing); curriculum effects (prev. episodes influence learning); and prioritized replay (surprising/reward events replayed more).
June 6, 2025 at 8:11 AM
Our framework explains a host of classic phenomena: DRM false memories (schema-consistent lures slip in when compressing); curriculum effects (prev. episodes influence learning); and prioritized replay (surprising/reward events replayed more).
Enter episodic memory as a life raft, which keeps surprising events in high fidelity, uncompressed by our current model. These vivid episodes rescue details needed to update semantic memory later when building the next version of the ship
June 6, 2025 at 8:11 AM
Enter episodic memory as a life raft, which keeps surprising events in high fidelity, uncompressed by our current model. These vivid episodes rescue details needed to update semantic memory later when building the next version of the ship