Unverified50% confidenceFactExact time
In tests on Qwen3.8 27B, the absolute KLD increase of the q4_0 KV cache harms more precise weights more: Q3-level weights increase by only 6-10%, UD-Q4_K_XL increases by about 39%, and Q8_0 increases by about 390%.
1
Sources
50%
Confidence
Long-term
Relevance
11/7/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe Q3_XXS model paired with an F16 KV cache scores approximately 80% on the needle test, while Q3_K_XL paired with a kvarn3 KV cache scores close to 100%.74% similarUnverifiedThe dense Qwen3.8-27B model's 3-bit version actually scored 0.05 higher (within noise range), but the number of tokens generated per task increased by 40% (confidence interval 1.25–1.59x).74% similarUnverifiedvLLM's FP8 KV cache adds 0.0004 to 0.0006 of KLD at every tested checkpoint on Qwen3.8 27B, a very small value that nonetheless appears consistently on almost every prompt.74% similarUnverifiedIn a comparison of Qwen3-0.6B in BF16 versus Q4_K_M, Q4_K_M changed the top-1 token at 11.6% of positions, with an average KL divergence of 0.067 and perplexity rising from 7.83 to 8.53.72% similarUnverifiedOn the Qwen3-VL-4B model, fine-tuning with procedurally generated data raises its accuracy on SpaceCast-Bench from 34.0% to 65.7%, and achieves a macro-average gain across six out-of-domain benchmarks.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1146818API
curl https://kongchang.com/api/v1/knowledge/claims/1146818MCP
get_claim(id=1146818)