[KongchangAI]
Unverified50% confidenceFactExact time

MAEM's training objective is label-free: the cosine similarity between the activation of the generated text, computed via a forward pass through the subject model, and the target direction vector serves as the reward signal, and it is optimized with reinforcement learning.

1
Sources
50%
Confidence
Long-term
Relevance
11/6/2026
First Seen

Sources

Related Entities

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/1142255
API
curl https://kongchang.com/api/v1/knowledge/claims/1142255
MCP
get_claim(id=1142255)