David Behery
@dahery.bsky.social
post-bac @kempnerinstitute.bsky.social at Harvard, studying intuitive physics in CCNLab under @gershbrain.bsky.social and @tomerullman.bsky.social
https://beheryd.github.io/
https://beheryd.github.io/
Reposted by David Behery
"abandoned teddy bears" category on wikimedia commons
December 6, 2025 at 12:14 AM
"abandoned teddy bears" category on wikimedia commons
Reposted by David Behery
Another fun project from @yangxiang.bsky.social. She asks the question: do people assign responsibility to personality traits in the same way that they assign reponsibility to people? The answer: sort of!
osf.io/preprints/ps...
osf.io/preprints/ps...
OSF
osf.io
December 6, 2025 at 3:11 PM
Another fun project from @yangxiang.bsky.social. She asks the question: do people assign responsibility to personality traits in the same way that they assign reponsibility to people? The answer: sort of!
osf.io/preprints/ps...
osf.io/preprints/ps...
Reposted by David Behery
Officially out in the current issue of Trends in Cognitive Sciences:
"Physics versus graphics as an organizing dichotomy in cognition"
www.cell.com/trends/cogni...
"Physics versus graphics as an organizing dichotomy in cognition"
www.cell.com/trends/cogni...
November 17, 2025 at 1:55 PM
Officially out in the current issue of Trends in Cognitive Sciences:
"Physics versus graphics as an organizing dichotomy in cognition"
www.cell.com/trends/cogni...
"Physics versus graphics as an organizing dichotomy in cognition"
www.cell.com/trends/cogni...
Reposted by David Behery
it's not that i ever doubted that being a barista was a skilled position, but my year-long effort to try to produce a half-decent latte at home has convinced me that a good barista is easily more skilled than virtually every c-suite executive on the planet
November 11, 2025 at 3:58 PM
it's not that i ever doubted that being a barista was a skilled position, but my year-long effort to try to produce a half-decent latte at home has convinced me that a good barista is easily more skilled than virtually every c-suite executive on the planet
Reposted by David Behery
overheard in coffee-shop convo now between two tech bros.:
"I don't want to think about AI 24/7, you know? I want to think about other stuff! Like...like...stoicism, you know?"
"I don't want to think about AI 24/7, you know? I want to think about other stuff! Like...like...stoicism, you know?"
October 24, 2025 at 5:38 PM
overheard in coffee-shop convo now between two tech bros.:
"I don't want to think about AI 24/7, you know? I want to think about other stuff! Like...like...stoicism, you know?"
"I don't want to think about AI 24/7, you know? I want to think about other stuff! Like...like...stoicism, you know?"
Reposted by David Behery
Reposted by David Behery
NEW on our #DeeperLearning blog
People balance being kind vs. being honest — and #LLMs should too.
New research shows training choices often favor informativeness over kindness, but prompting can induce sycophancy.
Read more: bit.ly/3Wqrtxl
People balance being kind vs. being honest — and #LLMs should too.
New research shows training choices often favor informativeness over kindness, but prompting can induce sycophancy.
Read more: bit.ly/3Wqrtxl
Using Cognitive Models to Reveal Value Trade-offs in Language Models - Kempner Institute
People’s actions and words are the result of a balance of different goals. The authors use a leading cognitive model of this value trade-off in polite speech to systematically examine […]
bit.ly
October 7, 2025 at 8:23 PM
NEW on our #DeeperLearning blog
People balance being kind vs. being honest — and #LLMs should too.
New research shows training choices often favor informativeness over kindness, but prompting can induce sycophancy.
Read more: bit.ly/3Wqrtxl
People balance being kind vs. being honest — and #LLMs should too.
New research shows training choices often favor informativeness over kindness, but prompting can induce sycophancy.
Read more: bit.ly/3Wqrtxl
Reposted by David Behery
#KempnerInstitute researchers find #ChatGPT’s guardrails are biased against certain users based on contextual information and inferences about the user.
Research by Victoria Li, @yidachen.bsky.social & @nsaphra.bsky.social.
Check out @theregister.com's article: bit.ly/47eaANa
#AI #LLMs
Research by Victoria Li, @yidachen.bsky.social & @nsaphra.bsky.social.
Check out @theregister.com's article: bit.ly/47eaANa
#AI #LLMs
ChatGPT hates LA Chargers fans
: Harvard researchers find model guardrails tailor query responses to user's inferred politics and other affiliations
bit.ly
August 28, 2025 at 4:45 PM
#KempnerInstitute researchers find #ChatGPT’s guardrails are biased against certain users based on contextual information and inferences about the user.
Research by Victoria Li, @yidachen.bsky.social & @nsaphra.bsky.social.
Check out @theregister.com's article: bit.ly/47eaANa
#AI #LLMs
Research by Victoria Li, @yidachen.bsky.social & @nsaphra.bsky.social.
Check out @theregister.com's article: bit.ly/47eaANa
#AI #LLMs
Wow! This is incredibly cool!
Our new paper, “A neural compass in the human brain during naturalistic human navigation” is out in @sfnjournals.bsky.social! First-author @zhenganglu.bsky.social led the charge, with Josh Julian and collaborator @gkaguirre.com.
www.jneurosci.org/content/earl...
www.jneurosci.org/content/earl...
August 21, 2025 at 1:32 AM
Wow! This is incredibly cool!
Reposted by David Behery
Applications are now open for our #KempnerInstitute Research Fellowships!
Postdocs studying the foundations of #intelligence or applications of #AI are encouraged to apply.
Learn more and apply by Oct. 1:
bit.ly/3UvpSpe
#LLMs #NeuroAI #ML @shamkakade.bsky.social
Postdocs studying the foundations of #intelligence or applications of #AI are encouraged to apply.
Learn more and apply by Oct. 1:
bit.ly/3UvpSpe
#LLMs #NeuroAI #ML @shamkakade.bsky.social
Kempner Research Fellowship - Kempner Institute
The Kempner brings leading, early-stage postdoctoral scientists to Harvard to work on projects that advance the fundamental understanding of intelligence.
bit.ly
August 15, 2025 at 1:00 PM
Applications are now open for our #KempnerInstitute Research Fellowships!
Postdocs studying the foundations of #intelligence or applications of #AI are encouraged to apply.
Learn more and apply by Oct. 1:
bit.ly/3UvpSpe
#LLMs #NeuroAI #ML @shamkakade.bsky.social
Postdocs studying the foundations of #intelligence or applications of #AI are encouraged to apply.
Learn more and apply by Oct. 1:
bit.ly/3UvpSpe
#LLMs #NeuroAI #ML @shamkakade.bsky.social
Reposted by David Behery
fun write up about our recent work on the limits of mental simulation:
news.harvard.edu/gazette/stor...
(original paper here: www.nature.com/articles/s41...)
news.harvard.edu/gazette/stor...
(original paper here: www.nature.com/articles/s41...)
Researchers uncover surprising limit on human imagination
Human beings can juggle up to 10 balls at once. But how many can they move through the air with their imaginations? The answer, published last month in Nature Communications,…
news.harvard.edu
August 13, 2025 at 7:05 PM
Reposted by David Behery
the Bayesian Models of Cognition book can teach you a lot about intuitive physics, e.g.:
August 2, 2025 at 3:56 PM
the Bayesian Models of Cognition book can teach you a lot about intuitive physics, e.g.:
Reposted by David Behery
Trans rights are human rights.
Fresh work incoming: Here we quantitatively compare the representational geometry of mouse CA1 and human spatial memory during environmental deformations. Briefest TLDR: Human memory looks like a change-resistant version of mouse CA1.
www.biorxiv.org/cgi/content/...
Fresh work incoming: Here we quantitatively compare the representational geometry of mouse CA1 and human spatial memory during environmental deformations. Briefest TLDR: Human memory looks like a change-resistant version of mouse CA1.
www.biorxiv.org/cgi/content/...
Change-resistance distinguishes the representational geometries of human spatial memory and mouse CA1 in deformed environments
Prior work has highlighted qualitative similarities between the neural instantiations of cognitive maps in rodents and memory-guided navigation in humans, suggesting a conservation of representational...
www.biorxiv.org
July 15, 2025 at 5:11 PM
Trans rights are human rights.
Fresh work incoming: Here we quantitatively compare the representational geometry of mouse CA1 and human spatial memory during environmental deformations. Briefest TLDR: Human memory looks like a change-resistant version of mouse CA1.
www.biorxiv.org/cgi/content/...
Fresh work incoming: Here we quantitatively compare the representational geometry of mouse CA1 and human spatial memory during environmental deformations. Briefest TLDR: Human memory looks like a change-resistant version of mouse CA1.
www.biorxiv.org/cgi/content/...
Reposted by David Behery
@cogscikid.bsky.social has cool new work, in collaboration with Honglak Lee and Sam Hall-McMaster, on how humans solve multitask reinforcement learning problems by preemptively simulating paths to counterfactual goals:
arxiv.org/abs/2507.05561
He also shows that it works on challenging ML tasks.
arxiv.org/abs/2507.05561
He also shows that it works on challenging ML tasks.
Preemptive Solving of Future Problems: Multitask Preplay in Humans and Machines
Humans can pursue a near-infinite variety of tasks, but typically can only pursue a small number at the same time. We hypothesize that humans leverage experience on one task to preemptively learn solu...
arxiv.org
July 10, 2025 at 9:52 AM
@cogscikid.bsky.social has cool new work, in collaboration with Honglak Lee and Sam Hall-McMaster, on how humans solve multitask reinforcement learning problems by preemptively simulating paths to counterfactual goals:
arxiv.org/abs/2507.05561
He also shows that it works on challenging ML tasks.
arxiv.org/abs/2507.05561
He also shows that it works on challenging ML tasks.
Glad to be here :)
The #KempnerInstitute is pleased to announce its four new 2025 Post-Baccalaureate Scholars! They will engage in research & coursework in preparation for applying to PhD programs in #intelligence research.
kempnerinstitute.harvard.edu/news/announc...
#neuroscience #NeuroAI #AI
kempnerinstitute.harvard.edu/news/announc...
#neuroscience #NeuroAI #AI
Announcing 2025 Kempner Post-Baccalaureate Scholars - Kempner Institute
Cambridge, MA– The Kempner Institute for the Study of Natural and Artificial Intelligence at Harvard is pleased to announce the 2025 cohort of scholars awarded placement in the Kempner Institute […]
kempnerinstitute.harvard.edu
July 4, 2025 at 12:56 PM
Glad to be here :)
Reposted by David Behery
it's the 10 year anniversary of a very important study
![In 2015, researchers Limb, Limb, Limb and Limb found that British doctors have surnames related to their specialties more often than expected by chance. Reproductive medicine had the largest proportion, with 1 in 52 doctors having specialty-relevant surnames including Horn, Hussey, and Woodcock. Second was Urology, with 1 in 59 related names such as Burns, Waterfall, Ball, and Koch.[96] "Dr Pain" appeared most commonly in general surgery.[96]](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:cyyd366g3hrxp5rynixm2mm3/bafkreiddq3sctguxn4xgf4mfxyu43qh7e2htev3iktz5siwagfpdswfy3q@jpeg)
June 13, 2025 at 12:10 PM
it's the 10 year anniversary of a very important study
Reposted by David Behery
My latest Aronov lab paper is now published @Nature!
When a chickadee looks at a distant location, the same place cells activate as if it were actually there 👁️
The hippocampus encodes where the bird is looking, AND what it expects to see next -- enabling spatial reasoning from afar
bit.ly/3HvWSum
When a chickadee looks at a distant location, the same place cells activate as if it were actually there 👁️
The hippocampus encodes where the bird is looking, AND what it expects to see next -- enabling spatial reasoning from afar
bit.ly/3HvWSum
June 11, 2025 at 10:24 PM
My latest Aronov lab paper is now published @Nature!
When a chickadee looks at a distant location, the same place cells activate as if it were actually there 👁️
The hippocampus encodes where the bird is looking, AND what it expects to see next -- enabling spatial reasoning from afar
bit.ly/3HvWSum
When a chickadee looks at a distant location, the same place cells activate as if it were actually there 👁️
The hippocampus encodes where the bird is looking, AND what it expects to see next -- enabling spatial reasoning from afar
bit.ly/3HvWSum
Reposted by David Behery
Out now in TiCS, something i've been thinking about a lot:
"Physics vs. graphics as an organizing dichotomy in cognition"
(by Balaban & me)
relevant for many people, related to imagination, intuitive physics, mental simulation, aphantasia, and more
authors.elsevier.com/a/1lBaC4sIRv...
"Physics vs. graphics as an organizing dichotomy in cognition"
(by Balaban & me)
relevant for many people, related to imagination, intuitive physics, mental simulation, aphantasia, and more
authors.elsevier.com/a/1lBaC4sIRv...
June 2, 2025 at 12:51 PM
Out now in TiCS, something i've been thinking about a lot:
"Physics vs. graphics as an organizing dichotomy in cognition"
(by Balaban & me)
relevant for many people, related to imagination, intuitive physics, mental simulation, aphantasia, and more
authors.elsevier.com/a/1lBaC4sIRv...
"Physics vs. graphics as an organizing dichotomy in cognition"
(by Balaban & me)
relevant for many people, related to imagination, intuitive physics, mental simulation, aphantasia, and more
authors.elsevier.com/a/1lBaC4sIRv...
Pizza in triangles? Pizza in squares? Nah… triangularly-latticed pizza.
June 1, 2025 at 10:06 PM
Pizza in triangles? Pizza in squares? Nah… triangularly-latticed pizza.
Reposted by David Behery
This framing is all wrong
Our international students are not a “crucial funding source”
They are our STUDENTS
They are the reason we EXIST
We teach STUDENTS
Our international students are not a “crucial funding source”
They are our STUDENTS
They are the reason we EXIST
We teach STUDENTS
May 22, 2025 at 6:21 PM
This framing is all wrong
Our international students are not a “crucial funding source”
They are our STUDENTS
They are the reason we EXIST
We teach STUDENTS
Our international students are not a “crucial funding source”
They are our STUDENTS
They are the reason we EXIST
We teach STUDENTS
Some of the best parts of my education so far have been centered around international collaboration. This is heartbreaking.
Breaking News: The Trump administration halted Harvard’s ability to enroll international students, a major escalation in its battle with the school.
Trump Administration Halts Harvard’s Ability to Enroll International Students
www.nytimes.com
May 22, 2025 at 7:14 PM
Some of the best parts of my education so far have been centered around international collaboration. This is heartbreaking.
“I didn’t forget your name… I just ended up in a local minima.”
Hopfield network in action: 4 patterns memorized, 1 recovered from noisy input.
Made with #python #numpy #matplotlib
Made with #python #numpy #matplotlib
May 16, 2025 at 7:28 PM
“I didn’t forget your name… I just ended up in a local minima.”
Reposted by David Behery
In a recent podcast @dyamins.bsky.social argues: "We want to be able to use the model of the digital brain as a neurobiomedical discovery and diagnostic design platform." neuroscience.stanford.edu/news/buildin... - That view is popular and I believe that there is a fundamental logical error. Thread
Building AI simulations of the human brain
In which Wu Tsai Neuro faculty scholar Dan Yamins explores how foundation models of the human brain could
neuroscience.stanford.edu
May 5, 2025 at 1:11 PM
In a recent podcast @dyamins.bsky.social argues: "We want to be able to use the model of the digital brain as a neurobiomedical discovery and diagnostic design platform." neuroscience.stanford.edu/news/buildin... - That view is popular and I believe that there is a fundamental logical error. Thread
