In the unique-hard-attention transformer model, chain-of-thought length must grow linearly with input size for parity, multiplication, median, and reachability.
Title resolution pending
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
-
Lower Bounds for Chain-of-Thought Reasoning in Hard-Attention Transformers
In the unique-hard-attention transformer model, chain-of-thought length must grow linearly with input size for parity, multiplication, median, and reachability.