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

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-GOMI-001
-category: "[[10_Wiki/💡 Topics/AI]]"
+category: "10_Wiki/💡 Topics/AI"
 confidence_score: 0.94
 tags: [auto-reinforced, goal-misgeneralization, ai-safety, alignment, reinforcement-learning, rewards, agent]
 last_reinforced: 2026-04-20
@@ -25,6 +25,6 @@ last_reinforced: 2026-04-20
 - **정책 변화(RL Update)**: 이제는 단순 보상 함수 정책(Reward function) 설계 정책을 넘어, 인간의 피드백 정책(RLHF)을 지속적으로 투입하여 AI 의 '내적 의도 정책'을 상시 검증하는 견제 정책(Anti-misgeneralization)이 안전 연구의 표준이 됨. (Effective-Altruism-in-AI와 연결)

 ## 🔗 지식 연결 (Graph)
- [[Reinforcement Learning (RL)]], [[Effective-Altruism-in-AI]], [[Bias-Variance-Tradeoff]], [[Reliability]], [[Safety]], [[Refinement]]
+- [[Reinforcement Learning (RL)]], [[Effective-Altruism-in-AI]], [[Bias-Variance-Tradeoff]], [[Reliability]], Safety, [[Refinement]]
 - **Key Researchers**: Lauro Langosco, Jack Koch et al.
 ---