The project EDAMOK (Enabling Distributed and Autonomous Management of Knowledge) aims at promoting a distributed approach to knowledge management, namely an approach based on the two following principles: (i) Principle of Autonomy: each organizational unit should be allowed a large degree of autonomy in managing (creating, representing, organizing, selecting, sharing) its own knowledge (โlocalโ knowledge); (ii) Principle of Coordination: knowledge sharing across organizational units should be thought of as a form of coordination between multiple autonomous perspectives rather than as a process of creating (and imposing) a supposedly shared knowledge structure. The goal of EDAMOK is to develop (i) a theoretical framework, a (ii) methodology, and (iii) a collection of technological tools to support this distributed and autonomous approach to knowledge management.
๐๏ธ 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.
๐Paper:
๐ปCode: https://github.com/lina-conti/voice-bias-coreference
#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?๐ง
๐๏ธ Our paper on connecting Speech Foundation Models with LLMs is featured in the SpeechLMM Training Journal on Weights & Biases.
Read it ๐ https://bit.ly/4svG7ll
SpeechLMM 2.0 coming this summer. ๐
#Meetween #SpeechLMM #AI #NLP