robertcprice

nCPU. nCPU: model-native and tensor-optimized CPU research runtimes with organized workloads, tools, and docs

entropy-seeding. Comprehensive analysis of PRNG vs TRNG vs QRNG entropy sources for Large Language Models across 7 model sizes (0.6B to 70B)

nQPU. Neural Quantum Processing Unit - High-performance quantum computing SDK with GPU acceleration

kvrm-selection-paper. Registry-Constrained Neural Selection for Verified System Interfaces

glm-mcp-server. GLM-4.7 MCP Server - Cost-efficient AI delegation for Claude Code with 87% savings

agent-monitor. Background daemon for monitoring AI agent sessions across Claude Code, Cursor, Aider, and custom agents

super-autoresearch. PyTorch MPS research platform for autonomous pretraining on Apple Silicon — warm-start studies, eval batch auto-probing, thermal-aware scheduling

KVRM. KVRM: registry-constrained decision architecture for audited finite-action routing — fail-closed action selection with versioned registries, support validation, and deterministic audit trails