Disfluency detection using a bidirectional LSTM V Zayats, M Ostendorf, H Hajishirzi arXiv preprint arXiv:1604.03209, 2016 | 131 | 2016 |
Audiopalm: A large language model that can speak and listen PK Rubenstein, C Asawaroengchai, DD Nguyen, A Bapna, Z Borsos, ... arXiv preprint arXiv:2306.12925, 2023 | 118 | 2023 |
Conversation modeling on Reddit using a graph-structured LSTM V Zayats, M Ostendorf Transactions of the Association for Computational Linguistics 6, 121-132, 2018 | 102 | 2018 |
Talking to the crowd: What do people react to in online discussions? A Jaech, V Zayats, H Fang, M Ostendorf, H Hajishirzi arXiv preprint arXiv:1507.02205, 2015 | 69 | 2015 |
Residual adapters for parameter-efficient ASR adaptation to atypical and accented speech K Tomanek, V Zayats, D Padfield, K Vaillancourt, F Biadsy arXiv preprint arXiv:2109.06952, 2021 | 51 | 2021 |
Disfluencies and human speech transcription errors V Zayats, T Tran, R Wright, C Mansfield, M Ostendorf arXiv preprint arXiv:1904.04398, 2019 | 43 | 2019 |
Giving attention to the unexpected: Using prosody innovations in disfluency detection V Zayats, M Ostendorf arXiv preprint arXiv:1904.04388, 2019 | 41 | 2019 |
Multi-domain disfluency and repair detection. V Zayats, M Ostendorf, H Hajishirzi INTERSPEECH, 2907-2911, 2014 | 40 | 2014 |
Representations for question answering from documents with tables and text V Zayats, K Toutanova, M Ostendorf arXiv preprint arXiv:2101.10573, 2021 | 38 | 2021 |
Exciton annihilation as bimolecular loss in organic solar cells L Tzabari, V Zayats, N Tessler Journal of Applied Physics 114 (15), 2013 | 36 | 2013 |
Disfluency detection with unlabeled data and small bert models JC Rocholl, V Zayats, DD Walker, NB Murad, A Schneider, DJ Liebling arXiv preprint arXiv:2104.10769, 2021 | 31 | 2021 |
Manipulating stance and involvement using collaborative tasks: an exploratory comparison. V Freeman, J Chan, GA Levow, RA Wright, M Ostendorf, V Zayats Interspeech, 303-307, 2014 | 19 | 2014 |
Robust cross-domain disfluency detection with pattern match networks V Zayats, M Ostendorf arXiv preprint arXiv:1811.07236, 2018 | 13 | 2018 |
Teaching BERT to wait: Balancing accuracy and latency for streaming disfluency detection A Chen, V Zayats, DD Walker, D Padfield arXiv preprint arXiv:2205.00620, 2022 | 12 | 2022 |
Unediting: Detecting disfluencies without careful transcripts V Zayats, M Ostendorf, H Hajishirzi Proceedings of the 2015 Conference of the North American Chapter of the
, 2015 | 6 | 2015 |
ATAROS Technical Report 1: Corpus collection and initial task validation V Freeman, J Chan, GA Levow, R Wright, M Ostendorf, V Zayats, Y Luan, ... U. Washington Linguistic Phonetics Lab, 2014 | 6 | 2014 |
Phonetic correlates of stance-taking V Freeman, R Wright, GA Levow, Y Luan, J Chan, T Tran, V Zayats, ... The Journal of the Acoustical Society of America 136 (4_Supplement), 2175-2175, 2014 | 5 | 2014 |
Robust preference optimization through reward model distillation A Fisch, J Eisenstein, V Zayats, A Agarwal, A Beirami, C Nagpal, P Shaw, ... arXiv preprint arXiv:2405.19316, 2024 | 4 | 2024 |
Multiturncleanup: A benchmark for multi-turn spoken conversational transcript cleanup H Shen, V Zayats, JC Rocholl, DD Walker, D Padfield arXiv preprint arXiv:2305.12029, 2023 | 4 | 2023 |
Zipper: A Multi-Tower Decoder Architecture for Fusing Modalities V Zayats, P Chen, M Merrari, D Padfield arXiv preprint arXiv:2405.18669, 2024 | | 2024 |