iiinova — where your digital self keeps evolving

6/ Check out the full paper in @cognitionjournal.bsky.social! 📄 authors.elsevier.com/a/1lMMD2Hx2-... 🧠 With @gershbrain.bsky.social and @annhuang42.bsky.social This study took almost 4 years of on-and-off work! The lesson learned here was to persevere on your ideas, even when you hit roadblocks! 🤩

authors.elsevier.com

261 6

5/ ✅ When memory is scarce, people chunk even more—they reduce their response time more when they have to remember more items (Ns=6). 💡𝗧𝗵𝗲 𝘁𝗮𝗸𝗲𝗮𝘄𝗮𝘆: chunking is an efficient and adaptive strategy to conserve cognitive resources.

130 3

4/ ✅ Our model also predicts (and our data confirm) that chunking in structured environments 𝗿𝗲𝗱𝘂𝗰𝗲𝘀 𝗽𝗼𝗹𝗶𝗰𝘆 𝗰𝗼𝗺𝗽𝗹𝗲𝘅𝗶𝘁𝘆 (the memory demands of the policy) while 𝗶𝗻𝗰𝗿𝗲𝗮𝘀𝗶𝗻𝗴 𝗲𝗮𝗿𝗻𝗲𝗱 𝗿𝗲𝘄𝗮𝗿𝗱.

120 2

3/ We tested the predictions of our model in behavioral experiments and found that: ✅ People chunk by leveraging temporal structure in the environment, making action selection 𝙡𝙚𝙨𝙨 𝙢𝙚𝙢𝙤𝙧𝙮-𝙞𝙣𝙩𝙚𝙣𝙨𝙞𝙫𝙚, 𝙛𝙖𝙨𝙩𝙚𝙧, 𝙖𝙣𝙙 𝙢𝙤𝙧𝙚 𝙖𝙘𝙘𝙪𝙧𝙖𝙩𝙚!

120 2

2/ We proposed that people learn and utilize action chunks because they are easier to store in memory. We formalized this idea in a model of 𝙘𝙤𝙣𝙙𝙞𝙩𝙞𝙤𝙣𝙖𝙡 𝙥𝙤𝙡𝙞𝙘𝙮 𝙘𝙤𝙢𝙥𝙧𝙚𝙨𝙨𝙞𝙤𝙣:

140 4

Lucy Lai @lucylai.bsky.social · 27/06/2025

🚨📊ɴᴇᴡ ᴘᴀᴘᴇʀ!! When we learn a new skill—like tying shoelaces or making pasta—we often start with a series of deliberate steps. But eventually, those steps blend into a smooth, single unit: an action chunk. But why and when do we chunk? 👇 📰https://authors.elsevier.com/a/1lMMD2Hx2-9B8

2539 53

I decided at age 5 that I wanted to be a teacher, and at 15 that I wanted to become a professor, so it feels surreal that this childhood dream is now a reality 🥹 I wrote a short reflection about how I got here, would love if you gave it a read! 📖: sabbatical.lucylai.com/p/becoming-p...

sabbatical.lucylai.com

becoming professor lai

fulfilling my childhood dreams

1292 29

Finally, if you are interested in teaching-focused faculty positions, please reach out to me anytime. I’ll also be sharing a guide on how to apply for such faculty positions very soon, so stay tuned! 🔔 #gotritons 6/6

050 5

These past two years in Japan 🇯🇵 have been so much more than I could have asked for. Thank you @oistedu.bsky.social for all the beautiful friendships made and for hosting me as a visiting scholar (when I wasn’t off exploring Asia 🌏✈️) 5/

100 1

I want to thank my mentors @gershbrain.bsky.social, @neurovenki.bsky.social, and @taralyntan.bsky.social for believing in my potential as a scientist and educator. And special shoutouts to @analog-ashley.bsky.social and Shannon Ellis for introducing me to teaching-focused faculty positions! 4/

120 2

Outside the classroom, my research will explore questions about 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻 𝗺𝗮𝗸𝗶𝗻𝗴, 𝗯𝗲𝗹𝗶𝗲𝗳 𝗰𝗵𝗮𝗻𝗴𝗲, 𝗶𝗻𝘁𝗲𝗹𝗹𝗲𝗰𝘁𝘂𝗮𝗹 𝗵𝘂𝗺𝗶𝗹𝗶𝘁𝘆, 𝗮𝗻𝗱 𝗵𝘂𝗺𝗮𝗻-𝗔𝗜 𝗰𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝗼𝗻 🤖🧠 3/

100 1

After a wonderful sabbatical year, I’m so excited to begin my career at UCSD doing what I love most: teaching students and making discoveries about human behavior and cognition 🧠 2/

100 1

Lucy Lai @lucylai.bsky.social · 30/04/2025

📢🚨I’m elated to share that I’ll be starting as a tenure-track 𝗔𝘀𝘀𝗶𝘀𝘁𝗮𝗻𝘁 𝗧𝗲𝗮𝗰𝗵𝗶𝗻𝗴 𝗣𝗿𝗼𝗳𝗲𝘀𝘀𝗼𝗿 👩🏻‍🏫 in the Department of Cognitive Science at UC San Diego this July! ☀️ @ucsandiego.bsky.social 1/ #ucsd #newprofessor #womeninSTEM

8794 79

ok so i know that journals will probably never pay reviewers but can't they at least introduce some incentive structure where you get publication costs waived for every 3 reviews you do or something? maybe not feasible if everyone wants to submit to CNS but you can centralize the incentive system 🤔

391 9

Lucy Lai @lucylai.bsky.social · 14/04/2025

great new work by @liushuze.bsky.social looking at how time and memory costs interact in policy compression :)

020 2

Lucy Lai @lucylai.bsky.social · 03/03/2025

🚨📢 so excited to be hosting this Global Summit on Open Problems in AI this April in Tokyo! 🇯🇵 we have an exciting line up of speakers from academia and industry. please share broadly! www.algopreneurship.ai #AI #research #entrepreneurship #conference #japan #summit

algopreneurship.ai

Algorithmic Innovation & Entrepreneurship

0103 10

Reposted by Lucy Lai

Reed Orchinik @rorchinik.bsky.social · 01/02/2025

New WP! The illusory truth effect (repetition -> belief) is core to psych of beliefs, & thought to be a deep bias impacting misinfo, persuasion & advertising Why would cognition include such a flaw? We argue it is a rational adaptation to high-quality info environments 🧵1/

47430 74

Lucy Lai @lucylai.bsky.social · 06/12/2024

this is the travel grant that changed my life and career!!! i applied as a sophomore in college and it’s what introduced me to the field of computational neuroscience 😍 i highly recommend you share with your undergrads / apply yourself if you are one! happy to answer any questions about it

051 5

Reposted by Lucy Lai

Sam Gershman @gershbrain.bsky.social · 12/01/2024

This is really cool work!

072 7

Lucy Lai @lucylai.bsky.social · 11/01/2024

apply to the FREE Okinawa Computational Neuroscience Summer School where you can hang out with cool people (incl. me!) at the beach 🏖️☀️ while learning computational neuroscience through lectures, tutorials, and independent projects!!! Deadline: Jan 31 groups.oist.jp/ocnc

040 4

should I cite chatGPT if it truly helped me start and finish my most recent project in three months? 🤔

020 2

looks like the place to be 👀

010 1

Thread summary here!: bsky.app/profile/lucy...

030 3

8/8 Our study contributes to a deeper understanding of how humans adapt their decision-making strategies under cognitive resource constraints, and may have the potential to inform the design of choice environments that align with human decision biases! #ChoosingBetter #NeuroSkyence

050 5

Policy compression: An information bottleneck in action selection

7/8 These results are uniquely explained by our policy compression model, which balances two computational goals: maximizing reward and minimizing policy complexity, which we previously developed in www.sciencedirect.com/science/arti...

sciencedirect.com

The brain has evolved to produce a diversity of behaviors under stringent computational resource constraints. Given this limited capacity, how do biol…

121 2

6/8 This bias persists even when multiple optimal choices are available 💰💰💰, and intensifies under cognitive load and time pressure, as illustrated in the above example.

110 1

5/8 Across three tasks, we found that people exploit structure in the relationships between states, actions, and rewards to “compress” their policies. Compressed policies are systematically biased towards actions that were most frequently chosen in the past. ♻️

110 1

4/8 To understand how cognitive capacity limits affect human decision making 🤔, we conducted experiments where we manipulated factors such as the number of optimal choices and the time limit within which choices were made.

100 1

3/8 But factors such as time constraints or mental fatigue can often steer us towards past choices, such as choosing to eat the same thing for lunch every day.

100 1

2/8 🍔 It is well known that decision making taxes cognitive resources. For example, when choosing what to eat for lunch (the action), a decision-maker might consider the nutritional content as well as the price of a meal (information about the state).

100 1