2 |
An Information-Theoretic Characterization of Morphological Fusion ...
|
|
|
|
BASE
|
|
Show details
|
|
3 |
Sensitivity as a Complexity Measure for Sequence Classification Tasks ...
|
|
|
|
Abstract:
We introduce a theoretical framework for understanding and predicting the complexity of sequence classification tasks, using a novel extension of the theory of Boolean function sensitivity. The sensitivity of a function, given a distribution over input sequences, quantifies the number of disjoint subsets of the input sequence that can each be individually changed to change the output. We argue that standard sequence classification methods are biased towards learning low-sensitivity functions, so that tasks requiring high sensitivity are more difficult. To that end, we show analytically that simple lexical classifiers can only express functions of bounded sensitivity, and we show empirically that low-sensitivity functions are easier to learn for LSTMs. We then estimate sensitivity on 15 NLP tasks, finding that sensitivity is higher on challenging tasks collected in GLUE than on simple text classification tasks, and that sensitivity predicts the performance both of simple lexical classifiers and of vanilla ... : Accepted by TACL. This is a pre-MIT Press publication version ...
|
|
Keyword:
Computation and Language cs.CL; Computational Complexity cs.CC; FOS Computer and information sciences; Machine Learning cs.LG
|
|
URL: https://dx.doi.org/10.48550/arxiv.2104.10343 https://arxiv.org/abs/2104.10343
|
|
BASE
|
|
Hide details
|
|
4 |
Sensitivity as a Complexity Measure for Sequence Classification Tasks ...
|
|
|
|
BASE
|
|
Show details
|
|
5 |
Universals of word order reflect optimization of grammars for efficient communication.
|
|
|
|
In: Proceedings of the National Academy of Sciences of the United States of America, vol 117, iss 5 (2020)
|
|
BASE
|
|
Show details
|
|
6 |
Crosslinguistic Word Orders Enable an Efficient Tradeoff of Memory and Surprisal
|
|
|
|
In: Proceedings of the Society for Computation in Linguistics (2020)
|
|
BASE
|
|
Show details
|
|
7 |
Theoretical Limitations of Self-Attention in Neural Sequence Models
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 8, Pp 156-171 (2020) (2020)
|
|
BASE
|
|
Show details
|
|
8 |
Tabula nearly rasa: Probing the Linguistic Knowledge of Character-Level Neural Language Models Trained on Unsegmented Text ...
|
|
|
|
BASE
|
|
Show details
|
|
10 |
CoMeT: integrating different levels of linguistic modeling for meaning assessment
|
|
|
|
BASE
|
|
Show details
|
|
11 |
Haribhaṭṭa in Nepal : ten legends from his Jātakamālā and the anonymous Śākyasiṃhajātaka
|
|
Haribhaṭṭa; Hahn, Michael (Herausgeber) - Darin enthalten: Śākyasiṃhajātaka. - Tokyo : The International Institute for Buddhist Studies, 2007
|
|
Institut für Empirische Sprachwissenschaft
|
|
UB Frankfurt Linguistik
|
|
Show details
|
|
13 |
Prakrit stanzas in an early anthology of Sanskrit verses
|
|
|
|
In: Bulletin d'études indiennes. - Paris : AFES 11-12 (1993), 355-368
|
|
Institut für Empirische Sprachwissenschaft
|
|
UB Frankfurt Linguistik
|
|
Show details
|
|
14 |
Über den indirekten Beweis bei literaturhistorischen Fragestellungen
|
|
|
|
In: Wiener Zeitschrift für die Kunde Südasiens. - Wien : Verl. der Österr. Akad. der Wiss. 36 (1992), 91-103
|
|
Institut für Empirische Sprachwissenschaft
|
|
UB Frankfurt Linguistik
|
|
Show details
|
|
17 |
Haribhaṭṭa and Gopadatta : two authors in the succession of Āryaśūra ; on the rediscovery of parts of their Jātakamālās
|
|
|
|
Institut für Empirische Sprachwissenschaft
|
|
UB Frankfurt Linguistik
|
|
Show details
|
|
20 |
Tabula nearly rasa: probing the linguistic knowledge of character-level neural language models trained on unsegmented text
|
|
|
|
BASE
|
|
Show details
|
|
|
|