TypeSafe AI 被用于学术引用检查工具
用户将 @typesafeai 加入其引用检查器并给予好评,可在 PaperTrellis 免费试用,标签涉及 Jev 与医学 AI。 链接
帖子报道了 Decision Bench 第5天的结果:Microsoft-Decision-1 以 96.4% 的准确率领先,而 Jev 1.13 仅以一项检查之差紧随其后,并提及隐藏问题数量与噪声评论成本。
Decision Bench, day 5: Microsoft-Decision-1 takes the lead. • 96.4% accuracy across 1,734 checks, working alone • 7 real questions hidden, against 1 for Jev 1.13 • 56 noise comments left in, about $0.076 for the whole run It edges past Jev 1.13 by one check. Jev 1.13 still