A general routing framework lets each token take its own path through a network, matching dense baselines while enabling learned compute savings and interpretable path specialization.
Fineweb-edu: the finest collection of educational content, 2024.https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu
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
-
Towards Distributed Neural Architectures
A general routing framework lets each token take its own path through a network, matching dense baselines while enabling learned compute savings and interpretable path specialization.