> Scientists from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) [...] trained a language model to predict the relationship between two s

> Scientists from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) [...] trained a language model to predict the relationship between two sentences, based on context and semantic meaning, using a dataset with labels for text snippets detailing if a second phrase “entails”, “contradicts”, or is neutral with respect to the first one. Using this dataset — natural language inference — they found that the newly trained models were significantly less biased than other baselines, without any extra data, data editing, or additional training algorithms > Many estimates say that the CO2 emission of training a language model can be higher than the lifelong emission of a car. Running these large language models is also very expensive because of the amount of parameters and the computational resources they need. [... Our] logical language model ... is 500 times smaller than the state-of-the-art models, can be deployed locally, and with no human-annotated training samples for downstream tasks. Our model uses 1/400 the parameters compared with the largest language models, has better performance on some tasks, and significantly saves computation resources [...] With demonstrated robust zero-shot adaptation ability, the model can be directly deployed to different tasks with more fairness, privacy, and better speed ...I cannot find the paper from Hongyin Luo and James Glass, Logic Against Bias: Textual Entailment Mitigates Stereotypical Sentence Reasoning. It is mentioned at https://2023.eacl.org/program/accepted/ https://2023.eacl.org/program/accepted/ , but there seem to be no links.