[KongchangAI]
Unverified50% confidenceBenchmarkExact time

Using less than 25% of labeled data, DySENSE can generate labeled high-fidelity synthetic energy consumption data and improve downstream prediction performance by 8% mIoU.

1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/16/2026
First Seen
Valid until: 2/14/2027

Sources

Related Entities

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/1174137
API
curl https://kongchang.com/api/v1/knowledge/claims/1174137
MCP
get_claim(id=1174137)