LLM Drift 방지를 위한 3계층 품질 측정 및 CI Gate 구축
Evaluating LLM Output Quality In Production
Evaluating LLM Output Quality In Production
Eval-Driven Agent Development: How I Stopped Tuning Prompts on Vibes
Swift VSX Support, Biome Type Inference, Agent Guardrails
Human Context 기반의 Modeling과 Padded Rooms 설계를 통한 LLM Slop 방어 전략
I benchmarked OpenAI's new GPT-Realtime-Translate against four other live translation systems
You Don't Align an AI, You Align with It
"Return JSON only" doesn't force JSON. Here's what actually forces it.
Token Budget Negotiator
Five habits that separate the operator from the vibe-coder
Proposal: A Real Benchmark for Long-Term AI Memory Systems