Zobrazeno 1 - 10
of 8 072
pro vyhledávání: '"Ostendorf, A."'
Large language models (LLMs) have demonstrated self-improvement capabilities via feedback and refinement, but current small language models (SLMs) have had limited success in this area. Existing correction approaches often rely on distilling knowledg
Externí odkaz:
http://arxiv.org/abs/2410.18209
Autor:
Wu, Junkai, Fan, Xulin, Lu, Bo-Ru, Jiang, Xilin, Mesgarani, Nima, Hasegawa-Johnson, Mark, Ostendorf, Mari
In recent years, we have observed a rapid advancement in speech language models (SpeechLLMs), catching up with humans' listening and reasoning abilities. SpeechLLMs have demonstrated impressive spoken dialog question-answering (SQA) performance in be
Externí odkaz:
http://arxiv.org/abs/2409.04927
Autor:
Hu, Yushi, Shi, Weijia, Fu, Xingyu, Roth, Dan, Ostendorf, Mari, Zettlemoyer, Luke, Smith, Noah A, Krishna, Ranjay
Humans draw to facilitate reasoning: we draw auxiliary lines when solving geometry problems; we mark and circle when reasoning on maps; we use sketches to amplify our ideas and relieve our limited-capacity working memory. However, such actions are mi
Externí odkaz:
http://arxiv.org/abs/2406.09403
Autor:
Saetchnikov, Anton V., Tcherniavskaia, Elina A., Saetchnikov, Vladimir A., Ostendorf, Andreas
In water monitoring, environmental analysis, cell culture stability, and biomedical applications, precise pH control is demanded. Traditional methods like pH strips and meters have limitations: pH strips lack precision, while electrochemical meters,
Externí odkaz:
http://arxiv.org/abs/2403.15117
Transformer-based NLP models are powerful but have high computational costs that limit deployment. Finetuned encoder-decoder models are popular in specialized domains and can outperform larger more generalized decoder-only models, such as GPT-4. We i
Externí odkaz:
http://arxiv.org/abs/2403.13112
Large language models (LLMs) have revolutionized the landscape of Natural Language Processing systems, but are computationally expensive. To reduce the cost without sacrificing performance, previous studies have explored various approaches to harness
Externí odkaz:
http://arxiv.org/abs/2311.09758
Autor:
Lu, Bo-Ru, Haduong, Nikita, Lee, Chia-Hsuan, Wu, Zeqiu, Cheng, Hao, Koester, Paul, Utke, Jean, Yu, Tao, Smith, Noah A., Ostendorf, Mari
The capabilities of pretrained language models have opened opportunities to explore new application areas, but applications involving human-human interaction are limited by the fact that most data is protected from public release for privacy reasons.
Externí odkaz:
http://arxiv.org/abs/2307.07047
Autor:
Saetchnikov, Anton V., Tcherniavskaia, Elina A., Saetchnikov, Vladimir A., Ostendorf, Andreas
The per- and polyfluoroalkyl substances (PFAS) constitute a group of organofluorine chemicals treated as the emerging pollutants and currently are of particularly acute concern. These compounds have been employed intensively as surfactants over multi
Externí odkaz:
http://arxiv.org/abs/2306.08908
Publikováno v:
The 5th Clinical Natural Language Processing Workshop. At ACL 2023
This paper explores methods for extracting information from radiology reports that generalize across exam modalities to reduce requirements for annotated data. We demonstrate that multi-pass T5-based text-to-text generative models exhibit better gene
Externí odkaz:
http://arxiv.org/abs/2306.09544
Autor:
Wu, Zeqiu, Hu, Yushi, Shi, Weijia, Dziri, Nouha, Suhr, Alane, Ammanabrolu, Prithviraj, Smith, Noah A., Ostendorf, Mari, Hajishirzi, Hannaneh
Language models (LMs) often exhibit undesirable text generation behaviors, including generating false, toxic, or irrelevant outputs. Reinforcement learning from human feedback (RLHF) - where human preference judgments on LM outputs are transformed in
Externí odkaz:
http://arxiv.org/abs/2306.01693