doko (@dorotheakolossa) 's Twitter Profile
doko

@dorotheakolossa

used to tweet for fun & science. believing in humanity & life on planet earth (despite it all)
@[email protected] @[email protected]

ID: 351742154

calendar_today09-08-2011 17:42:00

498 Tweet

342 Takipçi

701 Takip Edilen

Edmund Lalor (@edmundlalor) 's Twitter Profile Photo

We found that Whisper’s later (more linguistic contextualized) layers better predict EEG than earlier (less contextualized acoustic) layers. The predictions were also better than those based on speech acoustics (envelope, spectrogram) and language (GPT2 word surprisal).

We found that Whisper’s later (more linguistic contextualized) layers better predict EEG than earlier (less contextualized acoustic) layers. The predictions were also better than those based on speech acoustics (envelope, spectrogram) and language (GPT2 word surprisal).
Katharina Kanzan (@kathakritzelt) 's Twitter Profile Photo

Kenia führt einen neuen nationalen Feiertag am 13. November ein, an dem überall im Land Bäume gepflanzt werden sollen. Das Land plant bis 2032 15 Milliarden neue Bäume anzupflanzen. 🌳🌴🌲 Es gibt zu viele negative News, deswegen poste ich jeden Tag eine #guteNewsdesTages.

Frank Hutter (@frankrhutter) 's Twitter Profile Photo

I like this new paper on DL vs trees at the #NeurIPS 2023 datasets and benchmark track: arxiv.org/abs/2305.02997. One of their findings: #TabPFN performs well across 98 datasets: it ties for 1st place with CatBoost in predictive performance but is 200x faster to train. This is

I like this new paper on DL vs trees at the #NeurIPS 2023 datasets and benchmark track: arxiv.org/abs/2305.02997. One of their findings: #TabPFN performs well across 98 datasets: it ties for 1st place with CatBoost in predictive performance but is 200x faster to train. This is
doko (@dorotheakolossa) 's Twitter Profile Photo

Auf zum #DigitalGipfel ... die erste Veranstaltung, die mich bei der Registrierung gefragt hat, ob ich mit Personenschutz anreise.

Auf zum #DigitalGipfel ... die erste Veranstaltung, die mich bei der Registrierung gefragt hat, ob ich mit Personenschutz anreise.
Martin Bauer (@martinmbauer) 's Twitter Profile Photo

If you want to define a continuous addition you get the Riemann integral This is what happens if you want to define a continuous *product* One of the weirdest and most satisfying integrals you've ever seen and why it's important for physics (a quite technical 🧵) 1/12

If you want to define a continuous addition you get the Riemann integral

This is what happens if you want to define a continuous *product* 

One of the weirdest and most satisfying integrals you've ever seen and why it's important for physics

(a quite technical 🧵) 1/12
doko (@dorotheakolossa) 's Twitter Profile Photo

Open #PhD or #Postdoc Position in our team in Berlin :) Needed experience: - Machine learning for speech & video data - Generative video models - Fundamentals of auditory science & neurophysiology - Signal processing for EEG, audio & video All details: jobs.tu-berlin.de/en/job-posting…

Open #PhD or #Postdoc Position in our team in Berlin :) 
Needed experience:
- Machine learning for speech & video data
- Generative video models 
- Fundamentals of auditory science & neurophysiology
- Signal processing for EEG, audio & video
All details: jobs.tu-berlin.de/en/job-posting…
Genius Thinking (@geniusgtx) 's Twitter Profile Photo

Goodhart's Law: When a measure becomes a goal, it stops being a good measure. I.g: Exams and standard admission shifted the focus of education. It's no longer about the students, but about grades and pay.

Goodhart's Law:

When a measure becomes a goal, it stops being a good measure.

I.g: Exams and standard admission shifted the focus of education.

It's no longer about the students, but about grades and pay.
Nick Cummins (@nickcummins41) 's Twitter Profile Photo

Submitting health-related work to INTERSPEECH 2025? Remember to select Topic 14.05, "Speech and Language in Health: From Remote Monitoring to Medical Conversations (Special Session)", in the submission system!

Submitting health-related work to <a href="/ISCAInterspeech/">INTERSPEECH 2025</a>? Remember to select Topic 14.05, "Speech and Language in Health: From Remote Monitoring to Medical Conversations (Special Session)", in the submission system!
doko (@dorotheakolossa) 's Twitter Profile Photo

Excellent to know this article is out, and to see the list of authors: science.org/doi/10.1126/sc… I hope the importance of AI #alignment will come to be seen more clearly and more pressingly, now.

Security & Privacy in Speech Communication (@spsc_sig) 's Twitter Profile Photo

📢 SPSC Webinar: "Whether When What": Detection, Localization and Diarization of Partially Spoofed Audio 🗓️ Wed. July 17th, 10am CEST, Lin Zhang from BUT Speech ✒️ Register here: spsc-sig.org/form/webinar-r… (yes, correct link!) It's free and open to everyone!

doko (@dorotheakolossa) 's Twitter Profile Photo

Open professorship at TU Berlin: "Control Systems" Have a look at the newly opened position in our faculty at TU Berlin: jobs.tu-berlin.de/en/job-posting… Potential exciting collaborations, e.g., in TU Berlin's Cluster of Excellence Science of Intelligence #machineintelligence #academicjobs

doko (@dorotheakolossa) 's Twitter Profile Photo

Wir suchen Doktorand:innen für unser junges Team TU Berlin : Machine Learning für Audio- & Biosignale, 5 Jahre mit spannender projektorientierter Lehre & Forschung an menschzentrierten, multimodalen Technologien: jobs.tu-berlin.de/stellenausschr…

Wir suchen Doktorand:innen für unser junges Team <a href="/TUBerlin/">TU Berlin</a>
: Machine Learning für Audio- &amp; Biosignale, 5 Jahre mit spannender projektorientierter Lehre &amp; Forschung an menschzentrierten, multimodalen Technologien: jobs.tu-berlin.de/stellenausschr…
doko (@dorotheakolossa) 's Twitter Profile Photo

Dear all, as much as I used to love this place, this is my last tweet. If you're interested keeping in touch, let's connect via @dorotheakolossa.bsky.social and/or on doko@mastodon.social So long!