T. Florian Jaeger

Faculty opening in computational neuroscience and cognition, including language, development, learning through the lifespan, etc.

The OSF repo also contains a vignette for the R library, as well as a walk-through of the bootstrap analyses we applied. Hopefully this will make it easy for others to apply this model to their own data! osf.io/tpwmv/. Feedback welcome.

New R library STM bsky.app/profile/hlplab… by Santiago Barreda implementing incremental vowel normalization & categorization (Nearey & Assmann's PSTM). New paper that describes & uses that STM to model vowel perception bsky.app/profile/hlplab…

(P)STM infers vowel & talker-specific normalization parameters from single observations, predicting listeners' perception far better than other common normalization models (incl. Lobanov or C-CuRE).

DL captures human speech perception both *qualitatively* & *quantitatively* (R2>96%) for over 400 combinations of exposure and test items. Yet, previous DL models fail to capture important limitations. Specifically, we find that DL seems to proceed by remixing prev experience 2/2

Very excited about this: putting distributional learning (DL) models of adaptive speech perception to a strong, informative test sciencedirect.com/science/articl… by Maryann Tan. We use Bayesian ideal observers & adapters to assess whether DL predicts rapid changes in speech perception 1/2

Great opportunity!

@mjpost I've asked myself surprisingly similar questions :)

It occurred to me to wonder if my habit of chugging directly from a liter bottle of grape juice while standing in front of the fridge at work is frowned upon.

@quarbby +1 Even (or especially?) as a PI, I found my time in industry super helpful. There is also lots to be learned about management, effective meeting structures, collaborative coding, etc. And for PIs, it can remind us what training mentees would actually benefit from!

Hot take: if PhD admission numbers are reduced, why not go have a stint in the industry first? I worked & it helped me financially & emotionally. Taught me to present my ideas, deal w people & rejection, crystalised the ideas I’m working on now & save money

@tallinzen That's exactly what I was thinking about when I saw your tweet. Eg it is hard to compare to countries with mostly public, federally or state-funded, research universities. I imagine, in those environments, the overhead is partly hidden in that gov-provided funding.

a part of it has to be that the federal government effectively subsidizes overhead for other freeloading sponsors that don't allow overhead (foundations, industry). would be curious to know if the funding amounts from these sources are high enough to explain it though.

obviously I think it's bad to slash overhead on existing grants, and I also think any savings from cutting overhead should go into increasing direct costs. but there is something to think about here, how can universities in other countries get by with much lower overhead costs?

🚨Exciting news! We now have the first-ever complete #EEGManyLabs replication. This large-scale multi-site study revisits a key debate in EEG & reinforcement learning. A thread! 🧵👇 📄 Full paper: doi.org/10.1016/j.cort…

@mjpost You are asking for too much. Turns out the inversion of an alarm clock is NP complete.

@_hlplab_ I am grateful to share a frailty so common that workarounds are built into my computer. Next on my wishlist is software to make me go to bed on time.

I'm not sure what it says about me or (I suspect) people in general, that providing a five-second, reversible delay on actually sending an email is such a useful feature, for everything from fixing typos to clearing up poorly-worded points to preventing escalations.

More than 40 percent of #postdocs leave academia. Those who landed a coveted faculty position were more likely to have had a highly cited paper, changed their research topic between their #PhD and postdoc, or moved abroad after receiving their doctorate. pnas.org/doi/10.1073/pn… @PNASNews -> nature.com/articles/d4158… #ScienceCareer

Please share: Postdoctoral position available in high-resolution retinal imaging and video eye-tracking at the University of Rochester. buff.ly/4ja5fKj @aplabUR @CVSUoR @URNeuroscience @RochesterOptics @FlaumEye

Feedback welcome! & see @_wbushong_ 's thesis for more on this topic, and her very cool computational simulations--presenting a new effort to better understand what can be inferred from the types of data collected in studies on subcat infomaintenance during speech perception. /n

For how long is information about past speech input available in short-term memory? What evidence would inform this question? What does (not) follow from previous work on this question? Delighted to see this work by @klintonbicknell @_wbushong_ in JML sciencedirect.com/science/articl…

But Bayesian mixed-effect analyses also identify a previously undocumented tendency in all 4 datasets--unexpected under all existing accounts. We present initial simulations that suggests that a combination of attentional lapses & ideal integration might explain the data. /3

We show existing evidence is compatible with ideal maintenance & integration of uncertainty, and derive a stronger test of that hypothesis. 2 re-analysis & 2 new experiments find that the ideal observer's predictions fit listeners' behavior better than previous proposals /2

We revisit the classic work by Connine and colleagues, and show why its results are often misinterpreted. Using an ideal observer framework, we derive what would be expected if listeners maintained and integrated subcategorical information beyond word boundaries. /1