Unverified50% confidenceBenchmarkExact time
At 16× compression, AdaMem's substring match improves by up to 3.2 percentage points (5.5%) over the uniform allocation baseline, with an average relative gain of 3.4%.
1
Sources
50%
Confidence
Long-term
Relevance
9/22/2026
First Seen
Sources
Related Entities
Related Claims
Unverified在完全相同的模型下，仅靠Harness策略优化，ARC-AGI-3得分从13.3%提升到38.3%，同时输出Token节省约六倍68% similarUnverifiedWhen the candidate token acceptance rate is high, speculative decoding can effectively output 2–4 tokens per forward pass, improving overall generation speed by 1.5x–3x.66% similarUnverifiedCompared to baseline methods, Dude achieves improvements of up to 22.8% in recall and precision.63% similarUnverifiedAcross the Qwen3 series, DSpark's average draft acceptance rate improved by 30.9%, 26.7%, and 30% respectively compared to EAGLE-3, and by 16%–18% compared to DeepFlash.63% similarUnverifiedUnder the same model (Claude 4.6 Opus), OCR's precision is 33.9% versus 7.23% for the general-purpose agent, representing an approximately 4.7× improvement in precision.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/940078API
curl https://kongchang.com/api/v1/knowledge/claims/940078MCP
get_claim(id=940078)