List your company

Long-Context Generation on PG-19 (30K context length)

4.75Throughput Speedup (micro)SpecPV-2K
1.71322.50163.294.0784Dec 2, 2025

Evaluation results

MethodLinks
SpecPV-2KBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=2K2025.124.753.46
SpecPV-4KBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=4K2025.124.573.43
SpecPV-8KBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=8K2025.124.233.42
EAGLE3-YARNBackbone=LLaMA3.1-8B-Instruct2025.122.823.38
TriForceBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=4K2025.122.652.61
TokenSwiftBackbone=LLaMA3.1-8B-Instruct, Partial KV Budget=4K2025.121.831.73