Pick of the week @fbk_mt
Super interesting paper on the limitations of quantization, demonstrating how post-training quantization scales poorly in data.
https://t.co/tBuGTL0Myi
⭐ For our #PickOfTheWeek, this paper explores an important question for modern speech AI:
🎙️ Which Evaluation for Which Speech Model?
👥 Authors: @Maureendss , @EeshanDhekane
Speech foundation models are evolving rapidly, but evaluation practices are still fragmented.
👉🏻 What is being evaluated
👉🏻 Which capabilities are required
👉🏻 What task/protocol constraints define the setup
One key takeaway is that benchmark choice can strongly shape conclusions about model performance and generalization.
🏝️ Also at #LREC2026, Palma de Mallorca!
@luisabentivogli presented our second paper.
"Phonetic-based Ranking for Improved Pseudo-Labeling in Low-Resource ASR"
📄 Paper: https://t.co/X7x431OkNB
Bravi!! 🎉
🏝️ Yesterday at #LREC2026, Palma de Mallorca!
@lina_conti presented "Voice, Bias, and Coreference: An Interpretability Study of Gender in Speech Translation" at the poster session.
📄Paper: https://t.co/DOKHeEdSw4
💻Code: https://t.co/agdakUJiXP
#SpeechTranslation#NLProc
How does the granularity of speech-text pairs impact SpeechLLM performance, and what is the optimal way to interleave tokens? Furthermore, what are the best practices for generating synthetic data to boost training?🧐
🎙️ Our paper on connecting Speech Foundation Models with LLMs is featured in the SpeechLMM Training Journal on Weights & Biases.
Read it 👉 https://t.co/996YPS97Q0
SpeechLMM 2.0 coming this summer. 👀
#Meetween#SpeechLMM#AI#NLP
Meetween is part of the organising committee of #IWSLT2026 — the premier conference on spoken language translation. The Shared Task Evaluation Period is open. Working on #speechtranslation, instruction following, or #modelcompression?
Get involved now! 🔗 https://t.co/GbfzpadZyq
4 papers accepted at #LREC -COLING 2026 in Palma de Mallorca 🎉
Sarcasm detection, reference prediction, gender bias in speech translation, and low-resource ASR.
Meet the team, 11–16 May 2026 👉 https://t.co/v9B5twV3Lu
Follow us for paper deep-dives in the coming weeks.
#NLP
Many languages are left behind by AI translation. Not for lack of need, but lack of data. Meetween is closing that gap. See us at #TAUS2026 in Rome, 3–5 June. #Meetween#MultilingualAI
⏰ 4 days left to apply!
🎓 2 PhD positions still open:
🎯 Human-Centred Evaluation Frameworks for Multilingual Technologies
✨ Multimedia Personalization with Multimodal Large Language Models
📅 Deadline: 15 May 2026
🔗 Full details: https://t.co/ZagIUM3UCo
🌍 @lina_conti and @luisabentivogli are heading to #LREC2026 in Palma! They'll present two papers:
📄 "Voice, Bias, and Coreference: An Interpretability Study of Gender in Speech Translation"
Paper link: https://t.co/DOKHeEdkGw