This account is no longer active. Find me on 🦋.
Europe
Joined June 2012
- Tweets330
- Following956
- Followers5.3K
- Likes542
Pinned Tweet
I have stopped using Twitter much a while ago, for the usual reasons. I am now dipping my toes in bluer waters.
Find me there.
Grzegorz Chrupała 🇪🇺🇺🇦 retweeted
📣 Call for Satellite Workshops for #Interspeech2025!
🌍 Plan your event to foster dialogue in speech & language research.
Deadline for ISCA endorsement: Jan 8, 2025
Increase your workshop’s visibility as an Interspeech satellite event! Learn more: interspeech2025.org/
ALT interspeech2025.org call for satellites deadline january 8 2025
Grzegorz Chrupała 🇪🇺🇺🇦 retweeted
📢 Call for Special Session Proposals for #Interspeech2025!
🗣️ Deadline: Nov 8, 2024
Explore interdisciplinary & emerging areas in speech communication
Especially welcome: topics on Fair & Inclusive Speech Science & Tech! 🌍
Learn more: interspeech2025.org/
ALT call for special sessions deadline november 8 2024
I wonder how Schmidhuber is taking it.
BREAKING NEWS
The Royal Swedish Academy of Sciences has decided to award the 2024 #NobelPrize in Physics to John J. Hopfield and Geoffrey E. Hinton “for foundational discoveries and inventions that enable machine learning with artificial neural networks.”
We'll be hosting Interspeech 2025 in Rotterdam. The first call is out: send us your challenge proposals.
👋 Call for Challenge Proposals for #Interspeech2025!
🗣️ Submit your innovative ideas by 21 Oct. We're looking for challenges that align with "Fair & Inclusive Speech Science & Tech." Let's push the boundaries of speech research together! 🌍
Details: interspeech2025.org/
ALT interspeech2025.org call for challenges deadline: october 21, 2024
Our latest paper is on disentanglement of representations in models of spoken language. ☟
Is it possible to *disentangle* the internal representations of deep learning models? To separate the hidden states of recent spoken language models into one vector specialized for transcription to text, and another that preserves task-relevant acoustic features? #NLProc
🧵(1/9)
Indeep consortium meeting in Groningen, with @AriannaBisazza @wzuidema @gsarti_ @hmohebbi75 @JumeletJ @linguisticshen @antske @iris_hx and several others.
projects.illc.uva.nl/indeep/
If you're at #NAACL2024, check out @linguisticshen's talk tomorrow (June 19). The talk is about lexical tone in spoken language models aclanthology.org/2024.naacl-…. Session "Speech recognition, text-to-speech and SLU", Don Alberto 2, 11:00-12:30.
Our paper on lexical tone in spoken language models is now available as a preprint at arxiv.org/abs/2403.16865
One of our findings is shown in this figure: fine-tuning an English vs Mandarin model on ASR has the opposite effect on encoding of Mandarin tones.
Sharing some great news today. My paper titled "Encoding of lexical tone in self-supervised models of spoken language" has been accepted to the main NAACL 2024 conference!
Many thanks to my co-authors @Phonologician @afraalishahi @AriannaBisazza @gchrupala
A brief 🧵below:
Grzegorz Chrupała 🇪🇺🇺🇦 retweeted
Our pick of the week by @DennisFucci: "Homophone Disambiguation Reveals Patterns of Context Mixing in Speech Transformers" by @hmohebbi75, @gchrupala, @afraalishahi, et al., EMNLP 2023.
#NLProc #NLP #EMNLP2023 #speech #transformer #contextmixing #homopohone #disambiguation
Interesting work on context mixing in speech models, analyzing syntactic cues ➡️ aclanthology.org/2023.emnlp-…
The results are further evidence that previously generated tokens should not be disregarded when explaining predictions of encoder-decoder models wrt speech input.
@fbk_mt
Grzegorz Chrupała 🇪🇺🇺🇦 retweeted
The moment for “Transformer-specific Interpretability” tutorial is very approaching!
Check out our detailed program & an extensive suggested reading list here: projects.illc.uva.nl/indeep/…
🗓️ Thursday, 21 March
⏱️ 14:00-17:30 pm
📍 Radisson room
#EACL2024 #NLProc #XAI
Our upcoming #EACL2024 tutorial “Transformer-specific Interpretability” will focus on the trending type of interpretability that makes use of specific features of transformers for understanding LLMs, and discuss their pros & cons!
More info: projects.illc.uva.nl/indeep/…
#NLProc #XAI
Grzegorz Chrupała 🇪🇺🇺🇦 retweeted
If you're at EACL and still unsure how to spend your Thursday afternoon, consider attending the Transformer-specific Interpretability workshop at 14:00!
Recommended for people with and without experience in interpretability, it will start with a gentle intro in interpretability!
Grzegorz Chrupała 🇪🇺🇺🇦 retweeted
Sharing some great news today. My paper titled "Encoding of lexical tone in self-supervised models of spoken language" has been accepted to the main NAACL 2024 conference!
Many thanks to my co-authors @Phonologician @afraalishahi @AriannaBisazza @gchrupala
A brief 🧵below:
Grzegorz Chrupała 🇪🇺🇺🇦 retweeted
Value Zeroing, a faithful approach for analyzing context mixing in Transformers, is now available on @InseqLib main branch for all @huggingface text generation models! 🔀
🔍Paper introducing VZ: aclanthology.org/2023.eacl-m…
🐛VZ in Inseq: tinyurl.com/inseq-vz
Hosein will be on the road; don't miss it if you around these places.
Hi @NS_online I'm afraid what you put on my discount card does not look like my real surname.
@yournameisvalid
Grzegorz Chrupała 🇪🇺🇺🇦 retweeted
I have 2 openings for my VIDI project on modeling low-res morphologically rich languages, taking inspiration from child language acquisition insights!
@GroNlp @univgroningen
📢 a Postdoc (apply by 15 Feb)
rug.nl/about-ug/work-with-us…
📢 a PhD (apply by 11 Mar)
rug.nl/about-ug/work-with-us…
Grzegorz Chrupała 🇪🇺🇺🇦 retweeted
Our upcoming #EACL2024 tutorial “Transformer-specific Interpretability” will focus on the trending type of interpretability that makes use of specific features of transformers for understanding LLMs, and discuss their pros & cons!
More info: projects.illc.uva.nl/indeep/…
#NLProc #XAI