Unverified50% confidenceBenchmarkExact time
Across eleven datasets spanning images, tabular data, and text, Phase-HDC reduces memory by 16–23× compared to standard float32 Adam and by 4–6× compared to 8-bit Adam.
1
Sources
50%
Confidence
Long-term
Relevance
10/9/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedPhase-HDC loses approximately 5 accuracy points on average relative to float32 Adam.74% similarUnverifiedIn experiments where all other conditions were held constant and only the update rule was replaced, Phase-HDC achieved the accuracy of 6-bit momentum Adam while reducing memory to one-third.72% similarUnverifiedAWS Aurora通过6副本4/6写入法定人数（跨3个可用区各2副本）实现持久性，声称年化数据丢失率低于10^-1263% similarUnverifiedAdam 优化器的内存开销是 SGD 的三倍，这一代价催生了 Adafactor、8-bit Adam 等低内存优化器变体63% similarVerifiedINT4 quantization compresses neural network weights from FP32 to a 4-bit integer representation, reducing model size by approximately 8 times.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/990627API
curl https://kongchang.com/api/v1/knowledge/claims/990627MCP
get_claim(id=990627)