Systematic generalization with edge transformers
Recent research suggests that systematic generalization in natural language understanding
remains a challenge for state-of-the-art neural models such as Transformers and Graph …
remains a challenge for state-of-the-art neural models such as Transformers and Graph …
Finding needles in a haystack: Sampling structurally-diverse training sets from synthetic data for compositional generalization
Modern semantic parsers suffer from two principal limitations. First, training requires
expensive collection of utterance-program pairs. Second, semantic parsers fail to generalize …
expensive collection of utterance-program pairs. Second, semantic parsers fail to generalize …
Inducing Transformer's Compositional Generalization Ability via Auxiliary Sequence Prediction Tasks
Systematic compositionality is an essential mechanism in human language, allowing the
recombination of known parts to create novel expressions. However, existing neural models …
recombination of known parts to create novel expressions. However, existing neural models …
Compositional generalization in multilingual semantic parsing over Wikidata
Semantic parsing (SP) allows humans to leverage vast knowledge resources through
natural interaction. However, parsers are mostly designed for and evaluated on English …
natural interaction. However, parsers are mostly designed for and evaluated on English …
Data factors for better compositional generalization
Recent diagnostic datasets on compositional generalization, such as SCAN (Lake and
Baroni, 2018) and COGS (Kim and Linzen, 2020), expose severe problems in models …
Baroni, 2018) and COGS (Kim and Linzen, 2020), expose severe problems in models …
Mutual exclusivity training and primitive augmentation to induce compositionality
Recent datasets expose the lack of the systematic generalization ability in standard
sequence-to-sequence models. In this work, we analyze this behavior of seq2seq models …
sequence-to-sequence models. In this work, we analyze this behavior of seq2seq models …
Transformer module networks for systematic generalization in visual question answering
Transformers achieve great performance on Visual Question Answering (VQA). However,
their systematic generalization capabilities, ie, handling novel combinations of known …
their systematic generalization capabilities, ie, handling novel combinations of known …
Understanding robust generalization in learning regular languages
A key feature of human intelligence is the ability to generalize beyond the training
distribution, for instance, parsing longer sentences than seen in the past. Currently, deep …
distribution, for instance, parsing longer sentences than seen in the past. Currently, deep …
Compositional generalization in dependency parsing
Compositionality--the ability to combine familiar units like words into novel phrases and
sentences--has been the focus of intense interest in artificial intelligence in recent years. To …
sentences--has been the focus of intense interest in artificial intelligence in recent years. To …
On Evaluating Multilingual Compositional Generalization with Translated Datasets
Z Wang, D Hershcovich - arxiv preprint arxiv:2306.11420, 2023 - arxiv.org
Compositional generalization allows efficient learning and human-like inductive biases.
Since most research investigating compositional generalization in NLP is done on English …
Since most research investigating compositional generalization in NLP is done on English …