specific method · filed under model architecture
IndexPool
Compresses groups of indexer key vectors using weighted pooling to reduce latency and memory.
- source
- 1
- model
- 1
- lab adopt it
- 1
- strongest
- core
How sources treat it
One count per evidence span, weakest treatment to strongest.
core 1
Documented in
Evidence
1 span quoted from the sources, strongest treatment first.
IndexPool compresses groups of indexer key vectors through weighted pooling to hold down latency and memory.
coremodel architecturein GLM-5.3-FlashZ.ai
Filed alongside
Other methods under model architecture :: token mixer :: sparse attention :: sparse attention indexer.
Compressed Sparse Attention 2Hierarchical Sparse IndexerIndexCacheIndexShareLightning IndexerReindex ModeReuse ModeDense Warm-up StageFine-grained token selectionIndex BranchIndexer WarmupReuse QSA index selection across speculative decoding stepsAverage poolingBlock-causal scoringCompressed lightweight indexerCross-stage shared-state management for attention reuseDetached indexer-input optimizationIndex Branch outputIndex Branch value headMQA indexerSingle-head index key