Alert button
Picture for Shane Steinert-Threlkeld

Shane Steinert-Threlkeld

Alert button

Embedding structure matters: Comparing methods to adapt multilingual vocabularies to new languages

Sep 09, 2023
C. M. Downey, Terra Blevins, Nora Goldfine, Shane Steinert-Threlkeld

Figure 1 for Embedding structure matters: Comparing methods to adapt multilingual vocabularies to new languages
Figure 2 for Embedding structure matters: Comparing methods to adapt multilingual vocabularies to new languages
Figure 3 for Embedding structure matters: Comparing methods to adapt multilingual vocabularies to new languages
Figure 4 for Embedding structure matters: Comparing methods to adapt multilingual vocabularies to new languages
Viaarxiv icon

Evaluating Transformer's Ability to Learn Mildly Context-Sensitive Languages

Sep 02, 2023
Shunjie Wang, Shane Steinert-Threlkeld

Figure 1 for Evaluating Transformer's Ability to Learn Mildly Context-Sensitive Languages
Figure 2 for Evaluating Transformer's Ability to Learn Mildly Context-Sensitive Languages
Figure 3 for Evaluating Transformer's Ability to Learn Mildly Context-Sensitive Languages
Figure 4 for Evaluating Transformer's Ability to Learn Mildly Context-Sensitive Languages
Viaarxiv icon

The Weighted Möbius Score: A Unified Framework for Feature Attribution

May 16, 2023
Yifan Jiang, Shane Steinert-Threlkeld

Figure 1 for The Weighted Möbius Score: A Unified Framework for Feature Attribution
Figure 2 for The Weighted Möbius Score: A Unified Framework for Feature Attribution
Figure 3 for The Weighted Möbius Score: A Unified Framework for Feature Attribution
Figure 4 for The Weighted Möbius Score: A Unified Framework for Feature Attribution
Viaarxiv icon

Probing for Understanding of English Verb Classes and Alternations in Large Pre-trained Language Models

Sep 11, 2022
David K. Yi, James V. Bruno, Jiayu Han, Peter Zukerman, Shane Steinert-Threlkeld

Figure 1 for Probing for Understanding of English Verb Classes and Alternations in Large Pre-trained Language Models
Figure 2 for Probing for Understanding of English Verb Classes and Alternations in Large Pre-trained Language Models
Figure 3 for Probing for Understanding of English Verb Classes and Alternations in Large Pre-trained Language Models
Figure 4 for Probing for Understanding of English Verb Classes and Alternations in Large Pre-trained Language Models
Viaarxiv icon

Testing Pre-trained Language Models' Understanding of Distributivity via Causal Mediation Analysis

Sep 11, 2022
Pangbo Ban, Yifan Jiang, Tianran Liu, Shane Steinert-Threlkeld

Figure 1 for Testing Pre-trained Language Models' Understanding of Distributivity via Causal Mediation Analysis
Figure 2 for Testing Pre-trained Language Models' Understanding of Distributivity via Causal Mediation Analysis
Figure 3 for Testing Pre-trained Language Models' Understanding of Distributivity via Causal Mediation Analysis
Figure 4 for Testing Pre-trained Language Models' Understanding of Distributivity via Causal Mediation Analysis
Viaarxiv icon

Learning to translate by learning to communicate

Jul 14, 2022
C. M. Downey, Leo Z. Liu, Xuhui Zhou, Shane Steinert-Threlkeld

Figure 1 for Learning to translate by learning to communicate
Figure 2 for Learning to translate by learning to communicate
Viaarxiv icon

Language Models Use Monotonicity to Assess NPI Licensing

May 28, 2021
Jaap Jumelet, Milica Denić, Jakub Szymanik, Dieuwke Hupkes, Shane Steinert-Threlkeld

Figure 1 for Language Models Use Monotonicity to Assess NPI Licensing
Figure 2 for Language Models Use Monotonicity to Assess NPI Licensing
Figure 3 for Language Models Use Monotonicity to Assess NPI Licensing
Figure 4 for Language Models Use Monotonicity to Assess NPI Licensing
Viaarxiv icon

A multilabel approach to morphosyntactic probing

Apr 17, 2021
Naomi Tachikawa Shapiro, Amandalynne Paullada, Shane Steinert-Threlkeld

Figure 1 for A multilabel approach to morphosyntactic probing
Figure 2 for A multilabel approach to morphosyntactic probing
Figure 3 for A multilabel approach to morphosyntactic probing
Figure 4 for A multilabel approach to morphosyntactic probing
Viaarxiv icon

A Masked Segmental Language Model for Unsupervised Natural Language Segmentation

Apr 16, 2021
C. M. Downey, Fei Xia, Gina-Anne Levow, Shane Steinert-Threlkeld

Figure 1 for A Masked Segmental Language Model for Unsupervised Natural Language Segmentation
Figure 2 for A Masked Segmental Language Model for Unsupervised Natural Language Segmentation
Figure 3 for A Masked Segmental Language Model for Unsupervised Natural Language Segmentation
Figure 4 for A Masked Segmental Language Model for Unsupervised Natural Language Segmentation
Viaarxiv icon

Bilingual alignment transfers to multilingual alignment for unsupervised parallel text mining

Apr 15, 2021
Chih-chan Tien, Shane Steinert-Threlkeld

Figure 1 for Bilingual alignment transfers to multilingual alignment for unsupervised parallel text mining
Figure 2 for Bilingual alignment transfers to multilingual alignment for unsupervised parallel text mining
Figure 3 for Bilingual alignment transfers to multilingual alignment for unsupervised parallel text mining
Figure 4 for Bilingual alignment transfers to multilingual alignment for unsupervised parallel text mining
Viaarxiv icon