activity
20182022
most citedFrom Zero to Hero: On the Limitations of Zero-Shot Cross-Lingual Transfer with Multilingual Transformers

34 citations · 35 across the 4 of their papers we have counts for

collaborators

8 papers

cs.CL20221 cited

Word Order Does Matter (And Shuffled Language Models Know It)

Vinit Ravishankar, Mostafa Abdou, Artur Kulmizev +1

Recent studies have shown that language models pretrained and/or fine-tuned on randomly permuted sentences exhibit competitive performance on GLUE, putting into question the import…

cs.CL2021

The Impact of Positional Encodings on Multilingual Compression

Vinit Ravishankar, Anders Søgaard

In order to preserve word-order information in a non-autoregressive setting, transformer architectures tend to include positional knowledge, by (for instance) adding positional enc…

cs.CL2021

Attention Can Reflect Syntactic Structure (If You Let It)

Vinit Ravishankar, Artur Kulmizev, Mostafa Abdou +2

Since the popularization of the Transformer as a general-purpose feature encoder for NLP, many studies have attempted to decode linguistic structure from its novel multi-head atten…

cs.CL2020

The Sensitivity of Language Models and Humans to Winograd Schema Perturbations

Mostafa Abdou, Vinit Ravishankar, Maria Barrett +3

Large-scale pretrained language models are the major driving force behind recent improvements in performance on the Winograd Schema Challenge, a widely employed test of common sens…

cs.CL202034 cited

From Zero to Hero: On the Limitations of Zero-Shot Cross-Lingual Transfer with Multilingual Transformers

Anne Lauscher, Vinit Ravishankar, Ivan Vulić +1

Massively multilingual transformers pretrained with language modeling objectives (e.g., mBERT, XLM-R) have become a de facto default transfer paradigm for zero-shot cross-lingual t…

cs.CL2020

Do Neural Language Models Show Preferences for Syntactic Formalisms?

Artur Kulmizev, Vinit Ravishankar, Mostafa Abdou +1

Recent work on the interpretability of deep neural language models has concluded that many properties of natural language syntax are encoded in their representational spaces. Howev…