List your company

Long-Context Generation on PG-19 (10K context length)

2.92Throughput Speedup (micro-averaged)SpecPV-4K
1.34961.75732.1652.5727Dec 2, 2025

Evaluation results

MethodLinks
SpecPV-4KBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=4K, KV Cache Offloading=disabled2025.122.923.29
SpecPV-2KBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=2K, KV Cache Offloading=disabled2025.122.883.3
SpecPV-8KBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=8K, KV Cache Offloading=disabled2025.122.533.34
EAGLE3-YARNBackbone=LLaMA3.1-8B-Instruct, KV Cache Offloading=disabled, Context Adaptation=YARN2025.122.493.38
TriForceBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=4K, KV Cache Offloading=disabled2025.121.722.62
TokenSwiftBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=4K, KV Cache Offloading=disabled2025.121.411.64