Why does surprisal from larger Transformer-based language models provide a poorer fit to human reading times? BD Oh, W Schuler Transactions of the Association for Computational Linguistics 11, 336-350, 2023 | 119 | 2023 |
Comparison of structural parsers and neural language models as surprisal estimators BD Oh, C Clark, W Schuler Frontiers in Artificial Intelligence 5, 777963, 2022 | 53 | 2022 |
Transformer-based language model surprisal predicts human reading times best with about two billion training tokens BD Oh, W Schuler Findings of the Association for Computational Linguistics: EMNLP 2023, 1915-1921, 2023 | 32* | 2023 |
Entropy- and distance-based predictors from GPT-2 attention patterns predict reading times over and above GPT-2 surprisal BD Oh, W Schuler Proceedings of the 2022 Conference on Empirical Methods in Natural Language …, 2022 | 21 | 2022 |
Surprisal estimators for human reading times need character models BD Oh, C Clark, W Schuler Proceedings of the 59th Annual Meeting of the Association for Computational …, 2021 | 20 | 2021 |
Modeling morphological learning, typology, and change: What can the neural sequence-to-sequence framework contribute? M Elsner, AD Sims, A Erdmann, A Hernandez, E Jaffe, L Jin, ... Journal of Language Modelling 7 (1), 53-98, 2019 | 19 | 2019 |
Frequency explains the inverse correlation of large language models' size, training data amount, and surprisal's fit to reading times BD Oh, S Yue, W Schuler Proceedings of the 18th Conference of the European Chapter of the …, 2024 | 13 | 2024 |
Leading whitespaces of language models' subword vocabulary pose a confound for calculating word probabilities BD Oh, W Schuler Proceedings of the 2024 Conference on Empirical Methods in Natural Language …, 2024 | 11 | 2024 |
Team Ohio State at CMCL 2021 shared task: Fine-tuned RoBERTa for eye-tracking data prediction BD Oh Proceedings of the Workshop on Cognitive Modeling and Computational …, 2021 | 6 | 2021 |
Character-based PCFG induction for modeling the syntactic acquisition of morphologically rich languages L Jin, BD Oh, W Schuler Findings of the Association for Computational Linguistics: EMNLP 2021, 4367-4378, 2021 | 5 | 2021 |
Contributions of propositional content and syntactic category information in sentence processing BD Oh, W Schuler Proceedings of the Workshop on Cognitive Modeling and Computational …, 2021 | 5* | 2021 |
THOMAS: The hegemonic OSU morphological analyzer using seq2seq BD Oh, P Maneriker, N Jiang Proceedings of the 16th Workshop on Computational Research in Phonetics …, 2019 | 5 | 2019 |
Token-wise decomposition of autoregressive language model hidden states for analyzing model predictions BD Oh, W Schuler Proceedings of the 61st Annual Meeting of the Association for Computational …, 2023 | 4 | 2023 |
Exploring English online research and comprehension strategies of Korean college students BD Oh PQDT-Global, 2018 | 3 | 2018 |
Linear recency bias during training improves Transformers' fit to reading times C Clark, BD Oh, W Schuler arXiv preprint arXiv:2409.11250, 2024 | 2 | 2024 |
Coreference-aware surprisal predicts brain response E Jaffe, BD Oh, W Schuler Findings of the Association for Computational Linguistics: EMNLP 2021, 3351-3356, 2021 | 1 | 2021 |
Predicting L2 writing proficiency with computational indices based on N-grams BD Oh 외국어교육연구 (Foreign Language Education Research) 21, 1-20, 2017 | 1 | 2017 |
The impact of token granularity on the predictive power of language model surprisal BD Oh, W Schuler arXiv preprint arXiv:2412.11940, 2024 | | 2024 |
Empirical shortcomings of Transformer-based large language models as expectation-based models of human sentence processing BD Oh The Ohio State University, 2024 | | 2024 |