MIT Encyclopedia of Cognitive Sciences is over 25 years old by now and totally not up-to-date, but it is still immensely enjoyable to read.
However, if you want something newer and openaccess, try https://oecs.mit.edu/
(it sadly seems to lack some articles with relation to cultural anthropology)

How can we scale up the translation of ordinary sentences into formal #logic?

Bui et al. “introduce Text-JEPA (Text-based Joint-Embedding Predictive Architecture), …for converting natural language into first-order logic (NL2FOL).”

https://doi.org/10.48550/arXiv.2507.20491

Newsletter, my latest installment in the "how to think about the new AI models" series on Reasonable People

https://open.substack.com/pub/tomstafford/p/large-language-models-and-the-amazon

Are #languageModels vulnerable to #anchoring #bias?

Huang et al. generated the #SynAnchors dataset to find out.

Anchoring was more common in shallower layers of models.

A reflective reasoning strategy was usually most helpful.

https://doi.org/10.48550/arXiv.2505.15392

Encouraged by the reaction to yesterday's (very short) post : https://tomstafford.substack.com/p/ai-will-be-the-biro-of-thought

I have a set of similar pre-baked talking points on how to make sense of the AI/LLM revolution, so this may be the first in a series

Do smaller #languageModels learn to reason better from longer chains of thought?

Luo et al. found
- Larger CoT data subsets didn't increase reflective reasoning (Figure 3)
- Accuracy didn't beat baseline until ≅32k CoT tokens (Figure 2)

https://doi.org/10.48550/arXiv.2506.07712

What have papers about #cognitiveBias been focusing on since 2010?

Plenty of work on #behaviorChange or #visualization.

But human-computer interaction (#HCI) appeared in 2019 ...and became the plurality?

https://doi.org/10.1145/3706598.3713450

At what point does giving #AI examples of ideal reasoning fail to produce additional benefit?

With one popular #LLM analyzing clinical cases, improvement wasn't linear: accuracy peaked at around nine examples and then declined (on average).

https://doi.org/10.1093/ehjdh/ztaf068

Sad news: “BODEN — Professor Margaret (Maggie) Boden, renowned cognitive scientist and long-time member of the University of Sussex, died peacefully in Brighton on 18th July 2025, aged 88.”

https://www.theargus.co.uk/memorials/death-notices/death/30683058.margaret-maggie-boden/notice/

Mistaking meaningless claims like "Good #health lends subtle creativity to reality” for profound ideas is known as #bullshit receptivity.

Susceptibility to such #BS was increased a bit by forcing people to quickly accept their initial impulse.

https://doi.org/10.1080/20445911.2025.2517038

New on 'Stuff': Book Review - Animals, Robots, Gods \n
https://stuff.graves.cl/posts/2025-07-20_12_53-book-review-animals,-robots,-gods.html

#anthropology
#ethnography
#psychology
#neuralnetworks
#philosophy
#cogsci
#ethics
#moral
#book-review
#️️️️
#english

ARC-3, a sneak peek at the next-gen, interactive reasoning benchmark designed to illuminate the capability gap between today's AI and tomorrow's AGI.

Play First 3 Games
https://three.arcprize.org/

As with previous ARC tests, the actual games used for testing AI are kept secret. AI algorithms must learn the games on the spot.

There are no instructions. You must play the game to discover controls, rules, and goal.

Interactive Reasoning Benchmarks (IRBs) test for a broad scope of capabilities:

• Exploration
• Percept -> Plan → Action
• Memory
• Goal Acquisition
• Alignment

Game Design Constraints

• Easy for humans (can pick it up in <1 min of game play)
• Core Knowledge Priors (no language, trivia, cultural symbols)
• Should require no instructions to play
• Should be fun for humans and playable in 5-10 minutes
• Innovative and novel game mechanics encouraged (Hidden state, theory of mind, long term planning, navigating other agents, etc.)

This week I'm posting about presentations from two cool events (over on Twitter): https://x.com/byrd_nick/status/1943219893291164057

What are the events?
(1) The 1st Experimental Argument Analysis workshop
(2) The 5th European #ExperimentalPhilosophy #Conference

For the next couple days, I'm posting about talks and posters from the 2025 BioXPhi Summit in #Switzerland. Follow on #BlueSky: https://bsky.app/profile/byrdnick.com/post/3lsim7t6gq22t

The #conference website: https://ibmb.unibas.ch/en/public-outreach/projects-to-the-public/basel-oxford-nus-bioxphi-summit-2025/

#AI reasoning models may seem to reason reflectively when they say things like, "Let me rethink that".

But do these "reflective" phrases predict better reasoning performance?

Not in #Deepseek R1 Zero: https://doi.org/10.48550/arXiv.2503.20783

Finding Peter Putnam https://lobste.rs/s/usc3yp #cogsci #historical #philosophy #science
https://nautil.us/finding-peter-putnam-1218035/

Can task-switching hinder decisions?

Switching between a reflection test and a fluid #IQ test lowered optimal reflection test scores and completion compared to taking the tests separately (N = 80).

Bad news for #multitasking?

https://ianburbidge.com/wp-content/uploads/2024/05/ian-burbidge-masters-dissertation-1.pdf

A #nudge improves a decision environment.
A #boost improves a decision competency.

This paper argues against Sunstein's suggestion that boosts are thus educative or reflective (System 2) interventions.

https://doi.org/10.1007/s11299-025-00324-1

Accepted in Res Philosophica

"Reflective" thinking is rife in #cogSci and the #history of ideas.
But we lack a unified definition.
So I synthesized one.
Just 2 key factors.
Not just unifying, but useful!

Audiopaper: https://byrdnick.com/archives/28904/upon-reflection-ep-15-a-two-factor-explication-of-reflection

Preprint: https://osf.io/preprints/psyarxiv/d628j