Unverified50% confidenceBenchmarkExact time
The author used Gemini 3 Flash to generate audience-style questions for 120 video caption passages to form "question–passage" data pairs, and after less than three minutes of training, the top-1 hit rate improved from approximately two-thirds to approximately three-quarters.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/8/2026
First Seen
Valid until: 1/6/2027
Sources
Related Entities
Related Claims
UnverifiedGemini 3.7 Flash多模态长视频理解得分85.4%，长上下文记忆保持率高达97%75% similarUnverified实测将一个小时的YouTube视频交给Gemini，不到一分钟就能解析出核心要点和思维逻辑69% similarUnverifiedOn the LongVideoBench long-video understanding benchmark, Gemini 3.7 Flash with agentic mode enabled achieved a dual win of reduced token consumption and improved accuracy.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/989457API
curl https://kongchang.com/api/v1/knowledge/claims/989457MCP
get_claim(id=989457)