Unverified50% confidenceBenchmarkExact time
Google Research reported a performance improvement of up to +11.5 points on the WorkBuddy-Bench benchmark.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/20/2026
First Seen
Valid until: 1/18/2027
Sources
Related Entities
Related Claims
UnverifiedThe core goal of this upgrade is to make Spark more precise and accurate when handling real work tasks, especially in scenarios involving collaboration with Google Workspace.68% similarUnverifiedGoogle publicly disclosed its recruitment data analysis conclusion: four rounds of structured interviews can achieve a predictive accuracy of 86%.66% similarUnverifiedGoogle 在2005年引入质量评分（Quality Score）机制，考量广告相关性、着陆页体验和预期点击率三大因素64% similarUnverifiedChain-of-Thought 技术在 2022 年由谷歌研究团队发表的论文中被系统证明可显著提升复杂任务准确率64% similarUnverifiedBIG-Bench 是由 Google 主导、超过 450 位研究者协作构建的基准测试集，包含 200 多个任务62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1075770API
curl https://kongchang.com/api/v1/knowledge/claims/1075770MCP
get_claim(id=1075770)