ENTITY
time per output token
time per output token
PulseAugur coverage of time per output token — every cluster mentioning time per output token across labs, papers, and developer communities, ranked by signal.
Total · 30d
2
2 over 90d
Releases · 30d
0
0 over 90d
Papers · 30d
2
2 over 90d
TIER MIX · 90D
TOPICS
SENTIMENT · 30D
2 day(s) with sentiment data
RECENT · PAGE 1/1 · 2 TOTAL
-
New UBEP library slashes MoE model latency on superpods · 2 sources tracked
Researchers have developed UBEP (Unified-Bus Expert Parallelism), a new communication library designed to optimize Mixture-of-Experts (MoE) models on large-scale superpods. UBEP addresses key bottlenecks in MoE communic…
-
New cascade framework optimizes LLM serving costs with minimal accuracy loss
A new research paper introduces a two-stage cascaded framework designed to optimize the cost of serving large language models (LLMs) in production. The system first clusters incoming queries to route them to the most co…