https://aclanthology.org/2022.findings-emnlp.7.pdf

#tda #transformers #attention

Acceptability Judgements via Examining the Topology of Attention Maps

"The role of the attention mechanism in encoding linguistic knowledge has received special interest in NLP. However, the attention heads’ ability to judge the grammatical acceptability of a sentence has been underexplored. This paper approaches the paradigm of acceptability judgments with topological data analysis (TDA), showing that the topological properties of the attention graph can be efficiently exploited for two standard practices in linguistics: binary judgments and linguistic minimal pairs. Topological features enhance the BERTbased acceptability classifier scores by up to 0.24 Matthew’s correlation coefficient score on COLA in three languages (English, Italian, and Swedish). By revealing the topological discrepancy between attention graphs of minimal pairs, we achieve the human-level performance on the BLIMP benchmark, outperforming nine statistical and Transformer LM baselines. At the same time, TDA provides the foundation for analyzing the linguistic functions of attention heads and interpreting the correspondence between the graph features and grammatical phenomena. We publicly release the code and other materials used in the experiments."

#tda #ml #ripser

Нашёл я пример, как TDA-библу Ripser применяют к supervised learning в виде feature extractor. Поверх фич Рипса идёт PCA: https://persim.scikit-tda.org/en/latest/notebooks/Persistence%20Landscapes%20and%20Machine%20Learning.html

Правда, Model accuracy: 0.59 всего составила. На train-test они смогли натвикать до 63%. Использовали SVC.

Интересно, уделает ли его UMAPS на этом примере, и какой будет верность просто с PCA..?