Undesirable memorization in large language models: A survey
Published in Arxiv, 2022
In this survey we provide a multi-perspective view and understanding of memorization.
Download here
Publications
Docalog: Multi-document Dialogue System using Transformer-based Span Retrieval
Published in ACL Workshop, 2022
This paper discusses our proposed approach, Docalog, for the DialDoc-22 (MultiDoc2Dial) shared task. Docalog identifies the most relevant knowledge in the associated document, in a multi-document setting. Docalog, is a three-stage pipeline consisting of (1) a document retriever model (DR. TEIT), (2) an answer span prediction model, and (3) an ultimate span picker deciding on the most likely answer span, out of all predicted spans. In the test phase of MultiDoc2Dial 2022, Docalog achieved f1-scores of 36.07% and 28.44% and SacreBLEU scores of 23.70% and 20.52%, respectively on the MDD-SEEN and MDD-UNSEEN folds.
Recommended citation: Sayed Hesam Alavian, Ali Satvaty, Sadra Sabouri, Ehsaneddin Asgari, and Hossein Sameti. 2022. Docalog: Multi-document Dialogue System using Transformer-based Span Retrieval. In Proceedings of the Second DialDoc Workshop on Document-grounded Dialogue and Conversational Question Answering, pages 142–147, Dublin, Ireland. Association for Computational Linguistics. https://aclanthology.org/2022.dialdoc-1.16/
A Change of Heart: Improving Speech Emotion Recognition through Speech-to-Text Modality Conversion
Published in Journal 1, 2010
Our paper proposes a modality conversion approach from audio to text in order to improve speech emotion recognition performance on the MELD dataset, this is feeling speech in words.
Download here