UNITOR @ Sardistance2020: Combining Transformer-based Architectures and Transfer Learning for Robust Stance Detection

Abstract

The dominant sequence transduction models are based on complex recurrent or convolutional neural networks that include an encoder and a decoder. The best performing models also connect the encoder and decoder through an attention mechanism. We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely. Experiments on two machine translation tasks show these models to be superior in quality while being more parallelizable and requiring significantly less time to train. Our model achieves 28.4 BLEU on the WMT 2014 Englishto-German translation task, improving over the existing best results, including ensembles, by over 2 BLEU. On the WMT 2014 English-to-French translation task, our model establishes a new single-model state-of-the-art BLEU score of 41.0 after training for 3.5 days on eight GPUs, a small fraction of the training costs of the best models from the literature. * Equal contribution. Listing order is random. Jakob proposed replacing RNNs with self-attention and started the effort to evaluate this idea. Ashish, with Illia, designed and implemented the first Transformer models and has been crucially involved in every aspect of this work. Noam proposed scaled dot-product attention, multi-head attention and the parameter-free position representation and became the other person involved in nearly every detail. Niki designed, implemented, tuned and evaluated countless model variants in our original codebase and tensor2tensor. Llion also experimented with novel model variants, was responsible for our initial codebase, and efficient inference and visualizations. Lukasz and Aidan spent countless long days designing various parts of and implementing tensor2tensor, replacing our earlier codebase, greatly improving results and massively accelerating our research.† Work performed while at Google Brain. ‡ Work performed while at Google Research.

SSSD: Leveraging Pre-Trained Models and Semantic Search for Semi-Supervised Stance Detection

“…Works as (Giorgioni et al, 2020;Ferreira and Vlachos, 2019) have proposed Transformer-based architectures combined with data augmentation and fine-tuning. They trained specific sentence classifiers based on UmBERTo using auxiliary datasets from tasks like sentiment analysis, irony detection, and hate-speech detection.…”

“…Transfer learning (Zhang et al, 2020;Giorgioni et al, 2020) and unsupervised approaches (Darwish et al, 2020;Rashed et al, 2021;Wei et al, 2019) are promising directions for SD, but they still face challenges in achieving comparable results to supervised machine learning approaches, especially in highly polarized environments such as Twitter. This is due to the difficulty of detecting stances in a noisy and polarized platform such as Twitter, where people express their opinions in nuanced and complex ways.…”

FANCY: A Diagnostic Data-Set for NLI Models

“…(i) We analyze emerging trends in the use of inti et al, 2020;Sarti, 2020;Giorgioni et al, 2020). This language model has 12-layer, 768hidden, 12-heads, 110M parameters.…”