Unverified50% confidenceBenchmarkExact time
Latent Compaction achieves 96.7% accuracy on the RULER benchmark at a 32x compression rate, while KVZip+ achieves only 40.0%.
1
Sources
50%
Confidence
Long-term
Relevance
10/15/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedGGUF量化格式（如Q4_K_M）通过将权重精度从32位浮点压缩到4位整数，通常只带来约1-5%的性能损失71% similarUnverifiedDevelopers ultimately selected the variant with the lowest KL divergence (0.028), resulting in a 1.83% drop on GSM8K, no change on MMLU, and an improvement in HarmBench compliance rate from 30.8% to 100%66% similarUnverifiedINT4量化将原本由65536个不同值（FP16）表示的权重压缩到仅16个离散级别65% similarVerifiedQ4量化将每个权重参数从16位压缩到4位，理论上可将模型体积缩小约75%64% similarUnverifiedW4A4量化将权重和激活值均压缩至4位整数，理论上相比FP16可将显存占用降低至约1/464% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1037891API
curl https://kongchang.com/api/v1/knowledge/claims/1037891MCP
get_claim(id=1037891)