gkirtzou
commented
1 year ago Material proposed in the meeting:
- Safety on linguistics constraint (not LLOD): https://www.sciencedirect.com/science/article/abs/pii/S0952197623009995
- Linguistic Framework and LLM: https://aclanthology.org/2022.naacl-main.325.pdf
- Use WordNet in vector space visualisation for understanding LLM learning process: https://aclanthology.org/2023.acl-long.308/
- DRS as neural model/graph (starting from this, quite a few approaches): https://direct.mit.edu/tacl/article/doi/10.1162/tacl_a_00241/43461/Exploring-Neural-Methods-for-Parsing-Discourse
- DRS, more recently Pre-Trained Language-Meaning Models for Multilingual Parsing and Generation
- Book: https://www.iospress.com/catalog/books/neuro-symbolic-artificial-intelligence-the-state-of-the-art
- Journal: https://www.iospress.com/catalog/journals/neurosymbolic-artificial-intelligence
This issue is meant to survey potential applications of neurosymbolic learning in relation to LLOD