Keyless Attention eliminates the key projection in transformers via value-space routing, cutting KV cache size by 50% with competitive or superior performance on language modeling and zero-shot benchmarks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Keyless Attention: Value-Space Routing and Value-Only Caching for Efficient Transformers
Keyless Attention eliminates the key projection in transformers via value-space routing, cutting KV cache size by 50% with competitive or superior performance on language modeling and zero-shot benchmarks.