Carlos G. Correa
@cgcorrea.bsky.social
postdoc with Marcelo Mattar - https://carlos.correa.me/
One neat direction I've heard of is @psychboyh.bsky.social & Robert Wilson's work using LLMs to process think-aloud data. 2024.ccneuro.org/pdf/67_Paper...
2024.ccneuro.org
December 5, 2024 at 12:41 PM
One neat direction I've heard of is @psychboyh.bsky.social & Robert Wilson's work using LLMs to process think-aloud data. 2024.ccneuro.org/pdf/67_Paper...
Shout out to my incredible co-authors: Sophia Sanborn (@naturecomputes.bsky.social), Mark Ho (@markkho.bsky.social), Fred Callaway (@fredcallaway.bsky.social), Nathaniel Daw (@nathanieldaw.bsky.social), and Tom Griffiths (@cocoscilab.bsky.social).
December 3, 2024 at 3:37 PM
Shout out to my incredible co-authors: Sophia Sanborn (@naturecomputes.bsky.social), Mark Ho (@markkho.bsky.social), Fred Callaway (@fredcallaway.bsky.social), Nathaniel Daw (@nathanieldaw.bsky.social), and Tom Griffiths (@cocoscilab.bsky.social).
Human behavior is hierarchically structured. But what determines *which* hierarchies people use? In a preprint, we run an experiment where people create programs that correspond to hierarchies, finding that people prefer structures with more reuse.
arxiv.org/abs/2311.18644
1/7
arxiv.org/abs/2311.18644
1/7
December 3, 2024 at 3:37 PM
Lightbot was originally built by Danny Yaroslavski. We build on an open-source version by Laurent Haan (github.com/haan/Lightbot).
Animation in first post was inspired by M. C. Escher’s Ascending and Descending and the game Monument Valley.
7/7
Animation in first post was inspired by M. C. Escher’s Ascending and Descending and the game Monument Valley.
7/7
February 13, 2024 at 4:38 PM
Lightbot was originally built by Danny Yaroslavski. We build on an open-source version by Laurent Haan (github.com/haan/Lightbot).
Animation in first post was inspired by M. C. Escher’s Ascending and Descending and the game Monument Valley.
7/7
Animation in first post was inspired by M. C. Escher’s Ascending and Descending and the game Monument Valley.
7/7
Huge thanks to my incredible co-authors: Sophia Sanborn, @markkho.bsky.social, @fredcallaway.bsky.social, @nathanieldaw.bsky.social, & @cocoscilab.bsky.social. I am incredibly grateful for their insights and support!
6/7
6/7
February 13, 2024 at 4:36 PM
Huge thanks to my incredible co-authors: Sophia Sanborn, @markkho.bsky.social, @fredcallaway.bsky.social, @nathanieldaw.bsky.social, & @cocoscilab.bsky.social. I am incredibly grateful for their insights and support!
6/7
6/7
For more details, check out the paper!
Preprint: arxiv.org/abs/2311.18644
Playable demo: carlos.correa.me/cocosci-ligh...
Experiment code: github.com/cgc/cocosci-...
Analysis code: github.com/cgc/lightbot...
5/7
Preprint: arxiv.org/abs/2311.18644
Playable demo: carlos.correa.me/cocosci-ligh...
Experiment code: github.com/cgc/cocosci-...
Analysis code: github.com/cgc/lightbot...
5/7
February 13, 2024 at 4:35 PM
For more details, check out the paper!
Preprint: arxiv.org/abs/2311.18644
Playable demo: carlos.correa.me/cocosci-ligh...
Experiment code: github.com/cgc/cocosci-...
Analysis code: github.com/cgc/lightbot...
5/7
Preprint: arxiv.org/abs/2311.18644
Playable demo: carlos.correa.me/cocosci-ligh...
Experiment code: github.com/cgc/cocosci-...
Analysis code: github.com/cgc/lightbot...
5/7
We find that people have a bias towards reuse, beyond the reuse that naturally occurs when minimizing program length. Drawing from theories of word learning, we account for this by modeling participants' program-writing as if they were creating & using an action grammar.
4/7
4/7
February 13, 2024 at 4:33 PM
We find that people have a bias towards reuse, beyond the reuse that naturally occurs when minimizing program length. Drawing from theories of word learning, we account for this by modeling participants' program-writing as if they were creating & using an action grammar.
4/7
4/7
Importantly, a program isn't just a sequence of instructions. Participants can define and use parts of programs (called processes), making it possible to write shorter, more compact programs.
3/7
3/7
February 13, 2024 at 4:33 PM
Importantly, a program isn't just a sequence of instructions. Participants can define and use parts of programs (called processes), making it possible to write shorter, more compact programs.
3/7
3/7
We use a process-tracing paradigm where people create hierarchical plans. Based on the educational game Lightbot, research participants drag and drop instructions to write programs. Lightbot follows these instructions, with the goal of activating all lights.
2/7
2/7
February 13, 2024 at 4:32 PM
We use a process-tracing paradigm where people create hierarchical plans. Based on the educational game Lightbot, research participants drag and drop instructions to write programs. Lightbot follows these instructions, with the goal of activating all lights.
2/7
2/7
Note: This thread is a repost from X (formerly twitter): twitter.com/_cgcorrea/st... The thread over there has several animations, which are only stills here.
1.5/7
1.5/7
February 13, 2024 at 4:32 PM
Note: This thread is a repost from X (formerly twitter): twitter.com/_cgcorrea/st... The thread over there has several animations, which are only stills here.
1.5/7
1.5/7