🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

My paper on hierarchical plans is out in Cognition!🎉 tldr: We ask participants to generate hierarchical plans in a programming game. People prefer to reuse beyond what standard accounts predict, which we formalize as induction of a grammar over actions. authors.elsevier.com/a/1kBQr2Hx2xLNA

Think your RLHF-trained AI is aligned with your goals? ⚠️ We found that RLHF can induce significant misalignment when humans provide feedback by predicting future outcomes 🤔, creating incentives for LLM deception 😱 Introduce ✨RLHS (Hindsight Simulation)✨: By simulating

Excited to share that our paper is now out in PNASNews! 🎉 Check it out: pnas.org/doi/10.1073/pn… Code and data: github.com/baixuechunzi/l… Big shoutout to my amazing coauthors Angelina Wang @angelinawang.bsky.social Ilia Sucholutsky Griffiths Computational Cognitive Science Lab!

New pre-print! In this work, we explore the extent to which different inductive biases can be instantiated among disparate neural architectures, specifically Transformers, CNNs, MLPs, and LSTMs. Link: arxiv.org/abs/2502.20237 (1/4)

New preprint reveals that large language models blend two distinct representations of numbers -- as strings and as integers -- which can lead to some surprising errors. This work shows how methods from cognitive science can be useful for understanding AI systems.

Many studies suggest AI has achieved human-like performance on various cognitive tasks. But what is “human-like” performance? Our new paper conducted a human re-labeling of several popular AI benchmarks and found widespread biases and flaws in task and label designs. We make 5

Happy to share my first first-authored work at Griffiths Computational Cognitive Science Lab. Determining sameness or difference between objects is utterly trivial to humans, but surprisingly inaccessible to AI. Meta-learning can help neural networks overcome this barrier. Link: arxiv.org/abs/2503.23212 (1/5)

We are looking for a new lab manager, shared with the Concepts and Cognition Lab of Tania Lombrozo. Apply here: research-princeton.icims.com/jobs/20656/res…

New paper: Language models have “universal” concept representation – but can they capture cultural nuance? 🌏 If someone from Japan asks an LLM what color a pumpkin is, will it correctly say green (as they are in Japan)? Or does cultural nuance require more than just language?

Every ChatGPT query costs more energy than the entire life of a fruit fly.

🚨 New preprint alert! 🚨 Thrilled to share new research on teaching! Work supervised by Griffiths Computational Cognitive Science Lab, Yael Niv @yaelniv.bsky.social, and Mark Ho. This project asks: When do people teach by mentalizing vs with heuristics? 1/3 osf.io/preprints/osf/…

(1/11) Evolutionary biology offers powerful lens into Transformers learning dynamics! Two learning modes in Transformers (in-weights & in-context) mirror adaptive strategies in evolution. Crucially, environmental predictability shapes both systems similarly.

New preprint! In-context and in-weights learning are two interacting forms of plasticity, like genetic evolution and phenotypic plasticity. We use ideas from evolutionary biology to predict when neural networks will use each kind of learning.

This paper uses metalearning to distill a Bayesian prior into a set of initial weights for a neural network, providing a way to create networks with interpretable soft inductive biases. The resulting networks can learn just as quickly as a Bayesian model when applied to new data.

New preprint shows that training large language models to produce better chains of thought for predicting human decisions also results in them producing better psychological explanations.

In this new preprint we use methods from cognitive science to explore how well large language models make inferences from observations and construct interventions for understanding complex black-box systems that are analogous to those that scientists seek to understand

Video games are a powerful tool for assessing the inductive biases of AI systems, as they are engineered based on how humans perceive the world and pursue their goals. This new benchmark evaluates the ability of vision language models using some challenging classic video games.