HiveMind reduces LLM agent failures under API contention from 72-100% to 0-18% by using admission control, backpressure, and priority queuing in a zero-modification proxy.
SWE-bench: Can language models resolve real-world GitHub issues?
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2representative citing papers
citing papers explorer
-
HiveMind: OS-Inspired Scheduling for Concurrent LLM Agent Workloads
HiveMind reduces LLM agent failures under API contention from 72-100% to 0-18% by using admission control, backpressure, and priority queuing in a zero-modification proxy.
- Resilient Write: A Six-Layer Durable Write Surface for LLM Coding Agents