Fact Tracing in LMs: MIT & Google Dataset and Benchmark Track Learned Knowledge Back to the Training Data

In the new paper Tracing Knowledge in Language Models Back to the Training Data, a team from MIT CSAIL and Google Research proposes a benchmark for tracing language models’ assertions to the associated training data, aiming to establish a principled ground truth and mitigate high compute demands for large neural language model training.

ai artificial intelligence benchmark data dataset deep-neural-networks google information-retrieval knowledge language model machine learning machine learning & data science mit ml natural language processing nature language tech research technology tracing training training data