Systematic Generalization
62 papers with code • 0 benchmarks • 7 datasets
Benchmarks
These leaderboards are used to track progress in Systematic Generalization
Libraries
Use these libraries to find Systematic Generalization models and implementations
SCAN
GSCAN
Mathematics Dataset
PCFG SET
S2B
Most implemented papers
Learning to Execute Programs with Instruction Pointer Attention Graph Neural Networks
google-research/google-research
•
•
NeurIPS 2020
More practically, we evaluate these models on the task of learning to execute partial programs, as might arise if using the model as a heuristic function in program synthesis.
Learning from Task Descriptions
Typically, machine learning systems solve new tasks by training on thousands of examples.
The role of Disentanglement in Generalisation
mmrl/disent-and-gen
•
•
ICLR 2021
It is claimed that such representations should be able to capture the compositional structure of the world which can then be combined to produce novel representations.
Systematic Generalization in Neural Networks-based Multivariate Time Series Forecasting Models
Hritikbansal/MultivariateTimeSeries_Generalization
•
•
Systematic generalization aims to evaluate reasoning about novel combinations from known components, an intrinsic property of human cognition.
Modelling the development of counting with memory-augmented neural networks
lucidrains/esbn-transformer
•
•
Learning to count is an important example of the broader human capacity for systematic generalization, and the development of counting is often characterized by an inflection point when children rapidly acquire proficiency with the procedures that support this ability.
SyGNS: A Systematic Generalization Testbed Based on Natural Language Semantics
We also find that the generalization performance to unseen combinations is better when the form of meaning representations is simpler.
Structured Reordering for Modeling Latent Alignments in Sequence Transduction
berlino/tensor2struct-public
•
•
NeurIPS 2021
Despite success in many domains, neural models struggle in settings where train and test examples are drawn from different distributions.
Lexicon Learning for Few-Shot Neural Sequence Modeling
Sequence-to-sequence transduction is the core problem in language processing applications as diverse as semantic parsing, machine translation, and instruction following.
How Modular Should Neural Module Networks Be for Systematic Generalization?
vanessadamario/understanding_reasoning
•
•
NeurIPS 2021
Neural Module Networks (NMNs) aim at Visual Question Answering (VQA) via composition of modules that tackle a sub-task.
Generalization and Robustness Implications in Object-Centric Learning
addtt/object-centric-library
•
•
The idea behind object-centric representation learning is that natural scenes can better be modeled as compositions of objects and their relations as opposed to distributed representations.
