VALSE 💃 accepted @ACL 2022 Main

Our paper VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena was accepted at ACL 2022 Main.


Neural Natural Language Generation: Multilinguality, Multimodality, Controllability and Learning

A survey by members of the COST Action Multi3Generation: Multi-task, Multilingual, Multi-modal, to appear in JAIR


New publication from the Heidelberg NLP group

We announce a new publication at *SEM 2022 at NAACL, on
A Dynamic, Interpretable CheckList for Meaning-oriented NLG Metric Evaluation – through the Lens of Semantic Similarity Rating. Congratulations to all authors!


Three new publications at AACL, EMNLP and Eval4NLP

from our team and collaborators on:
SBERT studying Meaning Representations, New challenges for AMR parsing metrics, and Adapter-based Fine-tuning in V&L models.
Congratulations to all authors!



Eight new publications by the Heidelberg NLP group, collaborating researchers and CL students

We are delighted to announce eight papers on interpretation and reasoning in NLU and NLG using Commonsense Knowledge, on Multilingual AMR Parsing, and on Image and Language processsing. Congratulations to all authors!


Best paper award at DeeLIO 2021

Best Paper Award for Maria Becker, Siting Liang and Anette Frank in the DeeLIO Workshop @ NAACL 2021

for their paper:
Reconstructing Implicit Knowledge with Language Models


New publications in TACL and ArgMining @ EMNLP

Novel AMR Graph Metrics and an AMR Graph Similarity Benchmark
Explainable Argument Similarity Rating for better Argument Conclusion Generation.


Researchers of the NLP group win best paper award in the 8th ArgMining Workshop @EMNLP 2021

for their paper Explainable Unsupervised Argument Similarity Rating with Abstract Meaning Representation and Conclusion Generation.