Unverified50% confidenceBenchmarkExact time
In the experiment, vocabulary sizes of 2,000, 8,000, 16,000, and 32,000 corresponded to tokens per byte (fertility) of 0.4475, 0.3667, 0.3435, and 0.3275, respectively.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe 2,103-token vocabulary is directionally wrong: at the 10M scale, the 2k vocabulary has a bpb about 19% higher than the 8k one.63% similarUnverifiedVocabulary choice matters more than backbone size: 10M labels with an 8k vocabulary (8.0M total parameters, bpb 1.168) beats 50M labels with a 2k vocabulary (33.5M total parameters, bpb 1.436) — 4.2 times fewer parameters but 18.7% lower bpb.62% similarUnverifiedActual testing of two quantization models: IQ2_XXS achieves 48 tokens/s for short Q&A and 63 tokens/s for long text; IQ3_XXS achieves 30 tokens/s for short Q&A and 39 tokens/s for long text, with VRAM usage of approximately 16.44 GB, activating 10,138 experts—IQ2 is about 60% faster than IQ3.62% similarUnverifiedIn a pydantic-ai test, after 10 parallel tool calls (21 new blocks), the next request read 6,827 tokens from cache; after 11 calls (23 blocks), the read count was 0 and 7,534 tokens were written again.61% similarUnverifiedIn speculative sampling, when the acceptance rate of candidate tokens is high, the effective number of output tokens per forward pass can reach 2–4, thereby improving the overall generation speed by 1.5x–3x.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1019328API
curl https://kongchang.com/api/v1/knowledge/claims/1019328MCP
get_claim(id=1019328)