📄ArXiv AI•較早收集於 10h
ERM Fixes Causal Rung Collapse in LLMs
#research#llms#v1#causal-reasoning#ermllms
⚡ 30-Second TL;DR
有什麼變化
Formalizes rung collapse as lack of gradient for P(Y|do(X)) vs P(Y|X)
為什麼重要
Addresses core reasoning flaws in LLMs, enabling better generalization and steerability. Could prevent entrenchment in production models, improving reliability across domains. Inverse scaling in steerability highlights need for targeted causal fixes.
下一步行動
Prioritize whether this update affects your current workflow this week.
誰應關注：Researchers & Academics
關鍵要點
- •Formalizes rung collapse as lack of gradient for P(Y|do(X)) vs P(Y|X)
- •Introduces ERM with Physical Grounding Theorem and AGM postulates
- •Demonstrates 3.7% persistence in reasoning-enhanced models, ERM boosts recovery
📰
AI 週報
閱讀本週精選 AI 大事摘要 →
👉相關動態
AI 策展新聞聚合。所有內容版權歸原始發布者所有。
原始來源： ArXiv AI ↗
每週 AI 簡報
每週一封，可隨時退訂。