Tangram makes non-uniform KV cache compression practical for LLM serving with deterministic budget allocation, head group paging, and ahead-of-time load balancing, achieving up to 2.6x throughput gains.
Transformers are multi-state RNNs
4 Pith papers cite this work, alongside 8 external citations. Polarity classification is still indexing.
4
Pith papers citing it
8
external citations · OpenAlex
citation-role summary
other 1
citation-polarity summary
years
2026 4verdicts
UNVERDICTED 4roles
other 1polarities
unclear 1representative citing papers
VaSE improves KV cache eviction accuracy for reasoning models by over 4% versus prior eviction methods at 4x compression through value-magnitude protection and stochastic diversity.
Training transformers with KV sparsification during continued pretraining produces representations that admit better post-hoc KV cache compression, improving quality under memory budgets for long-context tasks.
GRKV applies global ridge regression to KV cache merging for span-based retention in long-context LLMs, claiming to be the only method that improves benchmark performance with minimal overhead.
citing papers explorer
-
Tangram: Unlocking Non-Uniform KV Cache for Efficient Multi-turn LLM Serving
Tangram makes non-uniform KV cache compression practical for LLM serving with deterministic budget allocation, head group paging, and ahead-of-time load balancing, achieving up to 2.6x throughput gains.
-
Value-Aware Stochastic KV Cache Eviction for Reasoning Models
VaSE improves KV cache eviction accuracy for reasoning models by over 4% versus prior eviction methods at 4x compression through value-magnitude protection and stochastic diversity.
-
Training Transformers for KV Cache Compressibility
Training transformers with KV sparsification during continued pretraining produces representations that admit better post-hoc KV cache compression, improving quality under memory budgets for long-context tasks.
-
GRKV: Global Regression for Training-Free KV Cache Compression in Long-Context LLMs
GRKV applies global ridge regression to KV cache merging for span-based retention in long-context LLMs, claiming to be the only method that improves benchmark performance with minimal overhead.