A context-tagged chunked RNN reaches optimal accuracy on a synthetic community-sequence task with a BPTT window of 1, where a naive RNN needs a window of 7.
Learning long-term dependencies with gradient descent is difficult
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Temporal Chunking Enhances Recognition of Implicit Sequential Patterns
A context-tagged chunked RNN reaches optimal accuracy on a synthetic community-sequence task with a BPTT window of 1, where a naive RNN needs a window of 7.