Unverified50% confidenceBenchmarkExact time
On the GSM8K dataset, with a block size of 16, the average matched prefix length between ProVe proposal blocks and AR decoding is 6.6 tokens.
1
Sources
50%
Confidence
Long-term
Relevance
10/21/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedProVe-lossless's speedup is limited by the degree of match between proposal blocks and AR decoding, averaging 6.6 matched tokens on GSM8K, leaving room for improvement.79% similarUnverifiedIn the GSM8K test (1319 questions, 8192-token limit), BF16 scored 1249, FP8 scored 1241, MXFP8 scored 1250, and NVFP4 scored 1238, with the three quantized formats matching BF16 within a ±3 point tolerance.68% similarUnverifiedAccording to Artificial Analysis testing, Grok 4.6 consumes approximately 38k tokens per task at the MXX High tier, while Grok 4.7 surges to 81k tokens.63% similarUnverifiedBased on measurements from more than 7,500 real API counts, the median call for claude-in-chrome returning a screenshot is 1,855 tokens, while the median call returning only text is 373 tokens, with screenshot calls at about 5 times text calls.61% similarUnverifiedJev's detection rate for disguised-spelling names was 13 of 25, while gpt-5-mini's was 25 of 25; adding leet characters or dots to clean names raised Jev's probability by an average of 0.16 (0.32 in English), while gpt-5-mini had zero false positives on 125 disguised clean names.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1079142API
curl https://kongchang.com/api/v1/knowledge/claims/1079142MCP
get_claim(id=1079142)