Pinned Loading

1. Understand Understand Public

   Python

2. HILSA HILSA Public

   Reproducible study: inference-time verification vs. adaptive compute on Qwen2.5-1.5B / GSM8K

   Python

3. StaticCore StaticCore Public

   An empirical study and evaluation harness characterizing KV-Cache admission control policies and mitigations in vLLM V1.

   Python

4. streaming-ttc-cache-coupling streaming-ttc-cache-coupling Public

   Empirical benchmarking harness mapping the boundary between compute saturation and PagedAttention KV-cache preemption cascades under streamed test-time compute scaling.

   Python