bergelsonlab
@bergelsonlab.bsky.social
Official account for the Bergelson Lab at Harvard, sporadically maintained by the PI:).
Just a lab, trying to figure out how babies learn language, somehow caught in the crosshairs of gov't admin battles.
Just a lab, trying to figure out how babies learn language, somehow caught in the crosshairs of gov't admin battles.
yeah, of the osf options help.osf.io/article/148-..., looks like "no license" is the best "Nothing can be done with the content without the license holder’s consent." i mean that's what i want, where 'nothing' means 'you can read it & build on it yourself or think ab the ideas if you want'
Licensing
This help guide provides an overview of licensing research data and materials and how to license your project, registration or preprint on OSF. COS does not pr
help.osf.io
January 13, 2026 at 4:57 PM
yeah, of the osf options help.osf.io/article/148-..., looks like "no license" is the best "Nothing can be done with the content without the license holder’s consent." i mean that's what i want, where 'nothing' means 'you can read it & build on it yourself or think ab the ideas if you want'
yeah i'm a bit wary when *Penguin Random House* is on the side of 'the little guy' aka the scientists trying to science or the creatives trying to creative. but yes some kind of overt DO NOT LLM this that was actually enforceable would be cool. but cf U.S.A, rules ain't shit w/o enforceability.
January 13, 2026 at 4:34 PM
yeah i'm a bit wary when *Penguin Random House* is on the side of 'the little guy' aka the scientists trying to science or the creatives trying to creative. but yes some kind of overt DO NOT LLM this that was actually enforceable would be cool. but cf U.S.A, rules ain't shit w/o enforceability.
& yes i realize putting anything on al gore's internet surely means its being fed into LLMs, whether a 'sandboxed' one where universities are in bed with the companies or the directly profiting kind that circumvent any guardrails or feint of intellectual integrity.
i hate it here.
i hate it here.
January 13, 2026 at 4:20 PM
& yes i realize putting anything on al gore's internet surely means its being fed into LLMs, whether a 'sandboxed' one where universities are in bed with the companies or the directly profiting kind that circumvent any guardrails or feint of intellectual integrity.
i hate it here.
i hate it here.
i see the gross parts of this uchicago study is well-trodden by @mehr.nz and co last week. so many things to be grumpy about but honestly, can you all please just STOP IT AND THINK OF NEW THINGS WITH YOUR BRAINS. kthnxbai. #PreprintsAreForHumans
January 13, 2026 at 4:20 PM
i see the gross parts of this uchicago study is well-trodden by @mehr.nz and co last week. so many things to be grumpy about but honestly, can you all please just STOP IT AND THINK OF NEW THINGS WITH YOUR BRAINS. kthnxbai. #PreprintsAreForHumans
Haha. Theory of mind in general, maybe knowledge access In particular? I once met with a psycholinguist I will not name who said “so in my 2007 paper, I’m sure you know the one”… Ma’am. I assure you I do not.
January 11, 2026 at 3:45 AM
Haha. Theory of mind in general, maybe knowledge access In particular? I once met with a psycholinguist I will not name who said “so in my 2007 paper, I’m sure you know the one”… Ma’am. I assure you I do not.
No silly goose Emily was clearly saying that infant (researchers) are prolific(ally bad at making jokes)
January 8, 2026 at 10:30 PM
No silly goose Emily was clearly saying that infant (researchers) are prolific(ally bad at making jokes)
(actual science article is aabbout 10 "gifted word learner" dogs, testing if they could learn 2 new toy words across conditions. they could! 10 'typical' dogs couldn't www.science.org/doi/10.1126/...); 'functionally parallel' doing a lot of work in abstract; actual text worded more thoughtfully)
Dogs with a large vocabulary of object labels learn new labels by overhearing like 1.5-year-old infants
Children as young as 18 months can acquire novel words by overhearing third-party interactions. Demonstrating similar learning processes in nonhuman species would indicate that the social-cognitive sk...
www.science.org
January 8, 2026 at 9:58 PM
(actual science article is aabbout 10 "gifted word learner" dogs, testing if they could learn 2 new toy words across conditions. they could! 10 'typical' dogs couldn't www.science.org/doi/10.1126/...); 'functionally parallel' doing a lot of work in abstract; actual text worded more thoughtfully)