Lina Conti

PhD Student

    Short bio

    Lina Conti is a PhD student at the Fondazione Bruno Kessler and the University of Trento (Italy), specializing in the interpretability of multilingual and multimodal foundation models for translation, with a focus on understanding gender bias in speech translation systems.

    She holds a master's degree in Computational Linguistics from Université Paris Cité (France) and bachelor's degrees in Computer Science and in Philosophy from Sorbonne University (France).

    Research topics


    • interpretability

    • XAI

    • speech translation

    • gender bias

    Publications

    1. Conti, Lina; Fucci, Dennis; Gaido, Marco; Negri, Matteo; Wisniewski, Guillaume; Bentivogli, Luisa,
      in «»,
      Proceedings of the 8th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP,
      Association for Computational Linguistics,
      vol. ,
      n. ,
      2025
      , pp. 398-
      414
    2. Conti, Lina; Wisniewski, Guillaume,
      Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing,
      2023
      , pp. 10362-
      10371

    ⭐ 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.

    🏝️ 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:
    💻Code: 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?🧐

    Load More