feat: complete wikification of War Commander batch 1&2 and final grey dot cleanup

This commit is contained in:
2026-04-27 18:58:22 +09:00
parent 3424166ea2
commit 6b86b0da4c
2706 changed files with 9074 additions and 7273 deletions
@@ -1,6 +1,6 @@
---
id: P-REINFORCE-AUTO-RLHF-001
category: "[[10_Wiki/💡 Topics/AI]]"
category: "10_Wiki/💡 Topics/AI"
confidence_score: 0.99
tags: [auto-reinforced, llm, reinforcement-learning, rlhf, ai-alignment]
last_reinforced: 2026-04-20
@@ -30,6 +30,6 @@ RLHF(Reinforcement Learning from Human Feedback)는 대규모 언어 모델(LLM)
- **정책 변화(RL Update)**: 인간 피드백 과정에서 발생하는 데이터 라벨러의 주관적 편향이 모델에 투영될 위험이 지적됨에 따라, 최근에는 'AI가 AI를 피드백'하는 RLAIF(AI Feedback)나 DPO(Direct Preference Optimization)와 같은 탈-인간 피드백 정책이 연구 표준으로 부상함.
## 🔗 지식 연결 (Graph)
- [[Reinforcement Learning (RL)]], [[Proximal Policy Optimization (PPO)]], [[Foundational Models]], [[Ethics & AI]], [[Ps-Reinforce]]
- [[Reinforcement Learning (RL)]], [[Proximal Policy Optimization (PPO)]], Foundational Models, [[Ethics & AI]], [[Ps-Reinforce]]
- **Modern Tech/Tools**: OpenAI InstructGPT, Anthropic Claude, Meta Llama-2/3 RLHF.
---