Unverified50% confidenceBenchmarkExact time
Compared with TextMAS, J-Comm's performance gains range from 1.29 to 8.59 percentage points while generating 19.2% to 74.8% fewer text tokens.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/26/2026
First Seen
Valid until: 1/24/2027
Sources
Related Entities
Related Claims
UnverifiedThe optimization effect of abbreviated instructions was 0, with characters reduced by 4.8% but tokens increased by 0.6%.61% similarUnverifiedUsing BM25 to route note bodies and the first three messages of a session raised the hit rate of the top four suggestions from 41% to 76%, and halved the token consumption per message.60% similarUnverifiedAdding an embedding-based semantic cache in front of the LLM, with a cosine similarity threshold set at 0.92 and returning cached results directly when the model is consistent, means about 30% of repeated user queries hit the cache, reducing overall spending by another about 20%.59% similarUnverifiedThis approach reduces token usage by 58.3% and improves inference speed by 1.95×, with no loss in accuracy.58% similarUnverifiedOn LongMemEval, CTWM achieved a 24.48% token reduction while maintaining overall accuracy at the same level.57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1098425API
curl https://kongchang.com/api/v1/knowledge/claims/1098425MCP
get_claim(id=1098425)