Unverified50% confidenceTradeoffExact time
Phase-HDC loses approximately 5 accuracy points on average relative to float32 Adam.
1
Sources
50%
Confidence
Long-term
Relevance
10/9/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn 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.77% similarUnverifiedAcross 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.74% similarVerifiedINT8 quantization loses approximately 0.5% accuracy while halving VRAM usage; INT4 quantization compresses VRAM to 1/4 of the original with an accuracy loss of approximately 1–3%.68% similarVerified4-bit quantization reduces model size to 1/8 of the original at the cost of approximately 1–3% accuracy loss.67% similarUnverifiedAn EMNLP study found that 4-bit models suffer an average retrieval loss of 23% at 128,000 tokens, with 70B LLaMA losing one-third of its score while QWEN of the same scale loses almost none.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/990628API
curl https://kongchang.com/api/v1/knowledge/claims/990628MCP
get_claim(id=990628)