Hacker News
- A Parameter-Free Classification Method with Compressors https://aclanthology.org/2023.findings-acl.426/ 8 comments
- On Language Models for Creoles [pdf] https://aclanthology.org/2021.conll-1.5.pdf 2 comments
- Small Language Models Are Also Few-Shot Learners https://aclanthology.org/2021.naacl-main.185/ 26 comments
Lobsters
- “Low-Resource” Text Classification: A Parameter-Free Classification Method with Compressors https://aclanthology.org/2023.findings-acl.426/ 2 comments ai , performance
- "Low-Resource” Text Classification: A Parameter-Free Classification Method with Compressors https://aclanthology.org/2023.findings-acl.426.pdf 3 comments ai , pdf
- A new study revealed a significant gap between AI- and human-level “understanding” of humor and why a cartoon is funny. The AI performance matching cartoon to caption was only 62% accurate, behind humans’ 94%. Comparing human- vs. AI-generated explanations, humans’ were preferred roughly 2-to-1. https://aclanthology.org/2023.acl-long.41/ 62 comments science
- “Low-Resource” Text Classification: A Parameter-Free Classification Method with Compressors https://aclanthology.org/2023.findings-acl.426.pdf 2 comments languagetechnology
- [R] Bert and XLNet: 85% Neurons Redundant, 92% Removable for Downstream Tasks https://aclanthology.org/2020.emnlp-main.398.pdf 37 comments machinelearning
- Language models with custom vocabularies? https://aclanthology.org/2020.clinicalnlp-1.17/ 2 comments languagetechnology
- Are languages in mBERT already "aligned" in vector space? https://aclanthology.org/2021.wmt-1.66.pdf 5 comments languagetechnology
- [D] Subword tokenization and large pretrained models https://aclanthology.org/2022.acl-long.450/ 2 comments machinelearning