Lovely work led by Dharsh Kumaran and his team at Google DeepMind shedding new light on the metacognitive profile of LLMs We show that LLM confidence estimates are both stubborn and brittle all at once Paper: www.nature.com/articles/s42...

Competing Biases underlie Overconfidence and Underconfidence in LLMs - Nature Machine Intelligence

Kumaran et al. show that large language model (LLM) confidence is shaped by two competing biases: a choice-supportive bias that inflates confidence in initial answers, and a systematic overweighting o...

nature.com

Excited that this paper on biscriptality in L2 reading is out, which leverages ENRO data (meco-read.com/about-enro/). Led by doctoral student Naima Mansuri and co-authored w/ great trainees (@estehr.bsky.social, @antonioiniesta.bsky.social, and others not here). www.cambridge.org/core/journal...

This sums up my post-Phd career, so far:

Luk, G., Hernández-Rivera, E., Tarín, K., Chen, D., Jang, M., & Titone, D. (2025). Math attitudes and verbal memory in multilingual younger adults. Canadian Journal of Experimental Psychology / Revue canadienne de psychologie expérimentale. doi.org/10.1037/cep0...

Suppose Tom loses control of his body, and his bodily motions then cause an accident. Would it be right to say “Tom caused the accident?” A new paper explores the role that language plays in these links between agency and causation! 🔗👇

Arthur has written a wonderful, clear, super helpful introduction to contractualism as a foundation for moral psychology. Perfect for getting up to speed, teaching, a desk reference, or a sunny day at the beach. wires.onlinelibrary.wiley.com/doi/10.1002/...

Contractualist Moral Cognition: From the Normative to the Descriptive at Three Levels of Analysis

Visual summary of the paper. Taking inspiration from the contractualist tradition in moral philosophy is helpful to better understand morality at three interrelated levels of analysis: Its evolutiona...

wires.onlinelibrary.wiley.com

🥳🥳 New paper in @nathumbehav.nature.com: “When development constricts our moral circle." Contrary to popular belief, younger kids may start out with broader moral circles than older ones. Check it out here 👉 rdcu.be/eoaSe w/ @mattiwilks.bsky.social @karrineldner.bsky.social & Lucius Caviola

When development constricts our moral circle

Nature Human Behaviour - Although many believe our moral circles expand with age, this Perspective discusses an early-emerging tendency to care for others.

rdcu.be

Excited to share Nigel Flower’s new MEG study on neural Sentence Superiority in parallel visual presentation. This work offers insights into how the brain detects & repairs word order errors, w early detection at ~200ms and repair mechanisms at starting at ~350ms doi.org/10.1523/JNEU...

The spatiotemporal dynamics of bottom-up and top-down processing during at-a-glance reading

Like all domains of cognition, language processing is affected by top-down knowledge. Classic evidence for this is missing blatant errors in the signal. In sentence comprehension, one instance is fail...

doi.org

I might have gone too click-happy with the following lists…I just joined an I’m already following 205 people 😂

I’m so happy this paper is finally published on #bilingualism and #self-evaluations! Couldn’t be more grateful for getting the opportunity to write about some of our ideas on why some of the self-reports we often use in #multilingualism are not as stable as we’d like to believe!

Debra Titone@dtitone.bsky.social · 2y ago

Doing #bilingualism or #multilingualism research and wondering which lang background measure to use? We examined if #language solidarity biased self-evaluation accuracy. doi.org/nrvz Bottom line: %use self-evals were less biased than value-laden (how good am I?) self-evals. %use wins, to us!

This is so cool! Fingers crossed ✈️🤞🏾

Debra Titone@dtitone.bsky.social · 2y ago

A bilingualism/multilingualusm starter pack! Reach out if you’d like me to add or remove you. go.bsky.app/C5Lhp1C