Robert Hawkins
@rdhawkins.bsky.social
asst prof @Stanford linguistics | director of social interaction lab 🌱 | bluskies about computational cognitive science & language
Reposted by Robert Hawkins
New perspective paper (w/ @sedaakbiyik.bsky.social, Joseph Outa, & @minjaek.bsky.social ) in @natrevpsychol.nature.com ⚽💭🧠👶 : www.nature.com/articles/s44...
How physical information is used to make sense of the psychological world - Nature Reviews Psychology
Reasoning about minds and reasoning about physical objects are governed by two distinct systems. In this Perspective, Liu et al. review research from developmental psychology and cognitive neuroscienc...
www.nature.com
November 24, 2025 at 11:14 PM
New perspective paper (w/ @sedaakbiyik.bsky.social, Joseph Outa, & @minjaek.bsky.social ) in @natrevpsychol.nature.com ⚽💭🧠👶 : www.nature.com/articles/s44...
Reposted by Robert Hawkins
I'm excited to announce that I have been asked to join the Mamdani Transition Team to work on community safety issues. A New Era for NYC.
November 24, 2025 at 8:23 PM
I'm excited to announce that I have been asked to join the Mamdani Transition Team to work on community safety issues. A New Era for NYC.
Reposted by Robert Hawkins
We have a new paper in Science Advances proposing a simple test for bias:
Is the same person treated differently when their race is perceived differently?
Specifically, we study: is the same driver likelier to be searched by police when they are perceived as Hispanic rather than white?
1/
Is the same person treated differently when their race is perceived differently?
Specifically, we study: is the same driver likelier to be searched by police when they are perceived as Hispanic rather than white?
1/
November 24, 2025 at 6:14 PM
We have a new paper in Science Advances proposing a simple test for bias:
Is the same person treated differently when their race is perceived differently?
Specifically, we study: is the same driver likelier to be searched by police when they are perceived as Hispanic rather than white?
1/
Is the same person treated differently when their race is perceived differently?
Specifically, we study: is the same driver likelier to be searched by police when they are perceived as Hispanic rather than white?
1/
Reposted by Robert Hawkins
📍Excited to share that our paper was selected as a Spotlight at #NeurIPS2025!
arxiv.org/pdf/2410.03972
It started from a question I kept running into:
When do RNNs trained on the same task converge/diverge in their solutions?
🧵⬇️
arxiv.org/pdf/2410.03972
It started from a question I kept running into:
When do RNNs trained on the same task converge/diverge in their solutions?
🧵⬇️
November 24, 2025 at 4:43 PM
📍Excited to share that our paper was selected as a Spotlight at #NeurIPS2025!
arxiv.org/pdf/2410.03972
It started from a question I kept running into:
When do RNNs trained on the same task converge/diverge in their solutions?
🧵⬇️
arxiv.org/pdf/2410.03972
It started from a question I kept running into:
When do RNNs trained on the same task converge/diverge in their solutions?
🧵⬇️
Reposted by Robert Hawkins
My collaborator @aditijh.bsky.social has some cool work on cell-type LDS :)
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
Disentangling the Roles of Distinct Cell Classes with Cell-Type Dynamical Systems
Latent dynamical systems have been widely used to characterize the dynamics of neural population activity in the brain. However, these models typically ignore the fact that the brain contains multiple...
www.biorxiv.org
November 23, 2025 at 6:30 PM
My collaborator @aditijh.bsky.social has some cool work on cell-type LDS :)
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
Reposted by Robert Hawkins
Tokenisers are a vital part of LLMs, but how hard is it to find an optimal one? 🤔 Considering arbitrarily large alphabets, prior work showed this is NP-hard. But what if we use bytes instead? Or unary strings like a, aa, aaa, ...? In our new paper, we show this is still hard, NP-hard!
November 20, 2025 at 3:05 PM
Tokenisers are a vital part of LLMs, but how hard is it to find an optimal one? 🤔 Considering arbitrarily large alphabets, prior work showed this is NP-hard. But what if we use bytes instead? Or unary strings like a, aa, aaa, ...? In our new paper, we show this is still hard, NP-hard!
Reposted by Robert Hawkins
we released Olmo 3! lot of exciting stuff but wanna focus on:
🐟Olmo 3 32B Base, the best fully-open base model to-date, near Qwen 2.5 & Gemma 3 on diverse evals
🐠Olmo 3 32B Think, first fully-open reasoning model approaching Qwen 3 levels
🐡12 training datasets corresp to different staged training
🐟Olmo 3 32B Base, the best fully-open base model to-date, near Qwen 2.5 & Gemma 3 on diverse evals
🐠Olmo 3 32B Think, first fully-open reasoning model approaching Qwen 3 levels
🐡12 training datasets corresp to different staged training
November 20, 2025 at 6:20 PM
we released Olmo 3! lot of exciting stuff but wanna focus on:
🐟Olmo 3 32B Base, the best fully-open base model to-date, near Qwen 2.5 & Gemma 3 on diverse evals
🐠Olmo 3 32B Think, first fully-open reasoning model approaching Qwen 3 levels
🐡12 training datasets corresp to different staged training
🐟Olmo 3 32B Base, the best fully-open base model to-date, near Qwen 2.5 & Gemma 3 on diverse evals
🐠Olmo 3 32B Think, first fully-open reasoning model approaching Qwen 3 levels
🐡12 training datasets corresp to different staged training
Reposted by Robert Hawkins
proud to share this work, led by the brilliant @ilinabg.bsky.social, now out in Nature! Ilina finds that speech-sound neural processing is VERY similar in a language you know and one you don't. differences only emerge at the level of word boundaries and learnt statistical structure 🧠✨
The human brain responds to the sounds of both familiar and unfamiliar languages in a similar way, according to research in Nature. The findings might guide future approaches to language learning and rehabilitation. go.nature.com/4ppvsHb #Neuroskyence 🧪
November 20, 2025 at 7:11 PM
proud to share this work, led by the brilliant @ilinabg.bsky.social, now out in Nature! Ilina finds that speech-sound neural processing is VERY similar in a language you know and one you don't. differences only emerge at the level of word boundaries and learnt statistical structure 🧠✨
Reposted by Robert Hawkins
aclanthology.org/2025.finding...
"annotators strongly took the LLM suggestions, significantly changing the label distribution compared to the baseline. We show that when these labels created with LLM assistance are used to evaluate LLM performance, reported model performance significantly increases"
"annotators strongly took the LLM suggestions, significantly changing the label distribution compared to the baseline. We show that when these labels created with LLM assistance are used to evaluate LLM performance, reported model performance significantly increases"
Just Put a Human in the Loop? Investigating LLM-Assisted Annotation for Subjective Tasks
Hope Schroeder, Deb Roy, Jad Kabbara. Findings of the Association for Computational Linguistics: ACL 2025. 2025.
aclanthology.org
November 20, 2025 at 11:06 PM
aclanthology.org/2025.finding...
"annotators strongly took the LLM suggestions, significantly changing the label distribution compared to the baseline. We show that when these labels created with LLM assistance are used to evaluate LLM performance, reported model performance significantly increases"
"annotators strongly took the LLM suggestions, significantly changing the label distribution compared to the baseline. We show that when these labels created with LLM assistance are used to evaluate LLM performance, reported model performance significantly increases"
Reposted by Robert Hawkins
New paper: Evidence of crossmodal compensation in the conceptual representation of blind people: while they share visual associations with sighted people, blind people more strongly associate concepts with touch direct.mit.edu/opmi/article...
November 20, 2025 at 8:52 AM
New paper: Evidence of crossmodal compensation in the conceptual representation of blind people: while they share visual associations with sighted people, blind people more strongly associate concepts with touch direct.mit.edu/opmi/article...
Reposted by Robert Hawkins
Our paper “The cost of thinking is similar between large reasoning models and humans” is now out in PNAS! 🤖🧠
w/ @fepdelia.bsky.social, @hopekean.bsky.social, @lampinen.bsky.social, and @evfedorenko.bsky.social
Link: www.pnas.org/doi/10.1073/... (1/6)
w/ @fepdelia.bsky.social, @hopekean.bsky.social, @lampinen.bsky.social, and @evfedorenko.bsky.social
Link: www.pnas.org/doi/10.1073/... (1/6)
PNAS
Proceedings of the National Academy of Sciences (PNAS), a peer reviewed journal of the National Academy of Sciences (NAS) - an authoritative source of high-impact, original research that broadly spans...
www.pnas.org
November 19, 2025 at 8:14 PM
Our paper “The cost of thinking is similar between large reasoning models and humans” is now out in PNAS! 🤖🧠
w/ @fepdelia.bsky.social, @hopekean.bsky.social, @lampinen.bsky.social, and @evfedorenko.bsky.social
Link: www.pnas.org/doi/10.1073/... (1/6)
w/ @fepdelia.bsky.social, @hopekean.bsky.social, @lampinen.bsky.social, and @evfedorenko.bsky.social
Link: www.pnas.org/doi/10.1073/... (1/6)
Reposted by Robert Hawkins
new paper by Sean Westwood:
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
November 18, 2025 at 7:16 PM
new paper by Sean Westwood:
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
Reposted by Robert Hawkins
My paper is out!
Computational modeling of error patterns during reward-based learning show evidence that habit learning (value free!) supplements working memory in 7 human data sets.
rdcu.be/eQjLN
Computational modeling of error patterns during reward-based learning show evidence that habit learning (value free!) supplements working memory in 7 human data sets.
rdcu.be/eQjLN
A habit and working memory model as an alternative account of human reward-based learning
Nature Human Behaviour - In this study, Collins proposes an alternative dual-process (working memory and habit) model of reinforcement learning in humans.
rdcu.be
November 17, 2025 at 5:18 PM
My paper is out!
Computational modeling of error patterns during reward-based learning show evidence that habit learning (value free!) supplements working memory in 7 human data sets.
rdcu.be/eQjLN
Computational modeling of error patterns during reward-based learning show evidence that habit learning (value free!) supplements working memory in 7 human data sets.
rdcu.be/eQjLN
Reposted by Robert Hawkins
🧠Our new preprint is out on PsyArXiv!
We study how getting more feedback (seeing what you could have earned) and facing gains vs losses change the way people choose between risky and safe options.
🖇️Link: doi.org/10.31234/osf...
It's a thread🧶:
We study how getting more feedback (seeing what you could have earned) and facing gains vs losses change the way people choose between risky and safe options.
🖇️Link: doi.org/10.31234/osf...
It's a thread🧶:
November 16, 2025 at 12:09 PM
🧠Our new preprint is out on PsyArXiv!
We study how getting more feedback (seeing what you could have earned) and facing gains vs losses change the way people choose between risky and safe options.
🖇️Link: doi.org/10.31234/osf...
It's a thread🧶:
We study how getting more feedback (seeing what you could have earned) and facing gains vs losses change the way people choose between risky and safe options.
🖇️Link: doi.org/10.31234/osf...
It's a thread🧶:
popsicle stick joke writers room >> laffy taffy joke writers room
November 18, 2025 at 5:15 AM
popsicle stick joke writers room >> laffy taffy joke writers room
Reposted by Robert Hawkins
I was honored to speak at Princeton’s symposium on The Physics of John Hopfield: Learning & Intelligence this week. I sketched out a perspective that ties together some of our recent work on ICL vs. parametric learning, and some possible links to hippocampal replay: 1/
November 15, 2025 at 8:56 PM
I was honored to speak at Princeton’s symposium on The Physics of John Hopfield: Learning & Intelligence this week. I sketched out a perspective that ties together some of our recent work on ICL vs. parametric learning, and some possible links to hippocampal replay: 1/
Reposted by Robert Hawkins
Every academic should join the @aaup.org national and their local AAUP chapter. If your uni doesn't have one, start one! Colorado academics, check out @aaupcsu.bsky.social! We're active, growing and looking to collaborate. Join the resistance on academia's front lines.
WE WON!!!
The US District Court for the Northern District of California has issued a preliminary injunction in AAUP et al v. Trump et al (the wall to wall union lawsuit challenging the administration’s unlawful use of Title VI to reshape the University of California system.
Read the order here:
The US District Court for the Northern District of California has issued a preliminary injunction in AAUP et al v. Trump et al (the wall to wall union lawsuit challenging the administration’s unlawful use of Title VI to reshape the University of California system.
Read the order here:
democracyforward.org
November 15, 2025 at 1:37 AM
Every academic should join the @aaup.org national and their local AAUP chapter. If your uni doesn't have one, start one! Colorado academics, check out @aaupcsu.bsky.social! We're active, growing and looking to collaborate. Join the resistance on academia's front lines.
Reposted by Robert Hawkins
BREAKING: In AAUP et al v. Trump (wall-to-wall union lawsuit challenging the administration’s unlawful use of TItle VI to reshape the University of California system), the faculty and staff of the UC system WON!!!
We were granted our preliminary injunction! @aaup.org
We were granted our preliminary injunction! @aaup.org
November 14, 2025 at 11:55 PM
BREAKING: In AAUP et al v. Trump (wall-to-wall union lawsuit challenging the administration’s unlawful use of TItle VI to reshape the University of California system), the faculty and staff of the UC system WON!!!
We were granted our preliminary injunction! @aaup.org
We were granted our preliminary injunction! @aaup.org
Reposted by Robert Hawkins
WE WON!!!
The US District Court for the Northern District of California has issued a preliminary injunction in AAUP et al v. Trump et al (the wall to wall union lawsuit challenging the administration’s unlawful use of Title VI to reshape the University of California system.
Read the order here:
The US District Court for the Northern District of California has issued a preliminary injunction in AAUP et al v. Trump et al (the wall to wall union lawsuit challenging the administration’s unlawful use of Title VI to reshape the University of California system.
Read the order here:
democracyforward.org
November 15, 2025 at 1:21 AM
WE WON!!!
The US District Court for the Northern District of California has issued a preliminary injunction in AAUP et al v. Trump et al (the wall to wall union lawsuit challenging the administration’s unlawful use of Title VI to reshape the University of California system.
Read the order here:
The US District Court for the Northern District of California has issued a preliminary injunction in AAUP et al v. Trump et al (the wall to wall union lawsuit challenging the administration’s unlawful use of Title VI to reshape the University of California system.
Read the order here:
Reposted by Robert Hawkins
Our latest hyperscanning study tracks how neural coupling unfolds during genuine conversation and how it differs from simply reading scripted dialogue: osf.io/preprints/ps...
OSF
osf.io
November 14, 2025 at 7:27 PM
Our latest hyperscanning study tracks how neural coupling unfolds during genuine conversation and how it differs from simply reading scripted dialogue: osf.io/preprints/ps...
Reposted by Robert Hawkins
I'm recruiting a grad student! My lab at NYU (psych dept) studies the computational basis of moral cognition and aims to build AI systems that are aligned with human values. Now admitting a PhD student for Fall 2026. Apps due 12/1. sites.google.com/corp/site/sy...
Sydney Levine - Open Positions
Graduate Students (PhD program)
I will be accepting PhD students through the NYU psychology department for the current application cycle (for admission in Fall 2026). My lab is joint between the Co...
sites.google.com
November 14, 2025 at 2:18 AM
I'm recruiting a grad student! My lab at NYU (psych dept) studies the computational basis of moral cognition and aims to build AI systems that are aligned with human values. Now admitting a PhD student for Fall 2026. Apps due 12/1. sites.google.com/corp/site/sy...
Reposted by Robert Hawkins
As of this morning, Starbucks workers across the country are officially ON STRIKE. And we're prepared for this to become the biggest and longest ULP strike in Starbucks history.
Say #NoContractNoCoffee with us: DON'T BUY STARBUCKS for the duration of our open-ended ULP strike! $SBUX
Say #NoContractNoCoffee with us: DON'T BUY STARBUCKS for the duration of our open-ended ULP strike! $SBUX
November 13, 2025 at 11:33 AM
As of this morning, Starbucks workers across the country are officially ON STRIKE. And we're prepared for this to become the biggest and longest ULP strike in Starbucks history.
Say #NoContractNoCoffee with us: DON'T BUY STARBUCKS for the duration of our open-ended ULP strike! $SBUX
Say #NoContractNoCoffee with us: DON'T BUY STARBUCKS for the duration of our open-ended ULP strike! $SBUX
Reposted by Robert Hawkins
Prohibited activities include advising Chinese graduate students. For reference, a recent Georgetown report estimated 16% of STEM graduate students in the US are Chinese nationals.
This would take out entire fields at the knees, which is perhaps the point.
This would take out entire fields at the knees, which is perhaps the point.
“The prohibited activities would include joint research, co-authorship on papers, and advising a foreign graduate student or postdoctoral fellow. The language is retroactive, meaning any interactions during the previous 5 years could make a scientist ineligible for future federal funding.”
U.S. Congress considers sweeping ban on Chinese collaborations
Researchers speak out against proposal that would bar funding for U.S. scientists working with Chinese partners or training Chinese students
www.science.org
November 14, 2025 at 2:52 AM
Prohibited activities include advising Chinese graduate students. For reference, a recent Georgetown report estimated 16% of STEM graduate students in the US are Chinese nationals.
This would take out entire fields at the knees, which is perhaps the point.
This would take out entire fields at the knees, which is perhaps the point.
this is really bad
“The prohibited activities would include joint research, co-authorship on papers, and advising a foreign graduate student or postdoctoral fellow. The language is retroactive, meaning any interactions during the previous 5 years could make a scientist ineligible for future federal funding.”
U.S. Congress considers sweeping ban on Chinese collaborations
Researchers speak out against proposal that would bar funding for U.S. scientists working with Chinese partners or training Chinese students
www.science.org
November 14, 2025 at 5:20 AM
this is really bad
Reposted by Robert Hawkins
chicago!
how can faculty make sense of the rapidly shifting economic and epistemic conditions for learning? how should faculty exercise governance and exert expertise over expensive partnerships between universities and technology companies?
join us for Theory At The Bargaining Table Dec 4!
how can faculty make sense of the rapidly shifting economic and epistemic conditions for learning? how should faculty exercise governance and exert expertise over expensive partnerships between universities and technology companies?
join us for Theory At The Bargaining Table Dec 4!
November 13, 2025 at 1:39 PM
chicago!
how can faculty make sense of the rapidly shifting economic and epistemic conditions for learning? how should faculty exercise governance and exert expertise over expensive partnerships between universities and technology companies?
join us for Theory At The Bargaining Table Dec 4!
how can faculty make sense of the rapidly shifting economic and epistemic conditions for learning? how should faculty exercise governance and exert expertise over expensive partnerships between universities and technology companies?
join us for Theory At The Bargaining Table Dec 4!