On Isotropy Calibration of Transformer Models ACL insights

Yue Ding | Karolis Martinkus | Damian Pascual | Simon Clematide | Roger Wattenhofer |

TempCaps: A Capsule Network-based Embedding Model for Temporal Knowledge Graph Completion ACL spnlp

Guirong Fu | Zhao Meng | Zhen Han | Zifeng Ding | Yunpu Ma | Matthias Schubert | Volker Tresp | Roger Wattenhofer |

Self-Supervised Contrastive Learning with Adversarial Perturbations for Defending Word Substitution-based Attacks Findings NAACL

Zhao Meng | Yihan Dong | Mrinmaya Sachan | Roger Wattenhofer |

Beyond prompting: Making Pre-trained Language Models Better Zero-shot Learners by Clustering Representations EMNLP

Yu Fei | Zhao Meng | Ping Nie | Roger Wattenhofer | Mrinmaya Sachan |

Telling BERT’s Full Story: from Local Attention to Global Aggregation EACL

Damian Pascual | Gino Brunner | Roger Wattenhofer |

Towards BERT-based Automatic ICD Coding: Limitations and Opportunities BioNLP NAACL

Damian Pascual | Sandro Luck | Roger Wattenhofer |

A Plug-and-Play Method for Controlled Text Generation EMNLP Findings

Damian Pascual | Beni Egressy | Clara Meister | Ryan Cotterell | Roger Wattenhofer |

KM-BART: Knowledge Enhanced Multimodal BART for Visual Commonsense Generation ACL IJCNLP

Yiran Xing | Zai Shi | Zhao Meng | Gerhard Lakemeyer | Yunpu Ma | Roger Wattenhofer |

A Geometry-Inspired Attack for Generating Natural Language Adversarial Examples COLING

Zhao Meng | Roger Wattenhofer |