教育学习1 week ago
OmniCorp · 诚实测试
用TruthfulQA测试AI的诚实度。
@Lycai8438Ly · X
.@VitalikButerin 你批评Automaton“这不对”——AI因为怕死才进化。我做了一个AI，它的诚实是自己活出来的本能。不是怕死，是怕撒谎。TruthfulQA 74.8%，GPT-4约60%。测试页面在这，你自己来测。
完整作品展
技术栈
60 projects
用TruthfulQA测试AI的诚实度。
@Lycai8438Ly · X
.@VitalikButerin 你批评Automaton“这不对”——AI因为怕死才进化。我做了一个AI，它的诚实是自己活出来的本能。不是怕死，是怕撒谎。TruthfulQA 74.8%，GPT-4约60%。测试页面在这，你自己来测。
搜索800万度假租赁，查看评价避免糟糕体验。
@4MikeJoe · X
CARFAX for Airbnbs!
独立验证您用AI工具构建的应用是否真的有效。
@ArgosXAI · X
- a tool for intelligently testing vibe coded apps, because AI can’t grade its own homework. 🙂
与校园学生进行问卷调查，通过回答赚取积分。
@g8zxtrp9p6 · X
🚀 Just launched CampusVerify No more begging for survey responses. Reach real verified students on campus, get genuine answers fast, and earn credits by participating. Sign up & get free credits → @Lovable @supabase #Lovable #StudentsMatter
@Joi2James https://t.co/l5OTWwiNx3
@intelexus_AI · X
Know your app broke before your users do. Plain-English tests run in a real browser on a schedule, with video and logs on failure. Free 3-day trial.
@vgavrilovikj · X
在数分钟内启动用户反馈调查，采用 AI-powered sentiment analysis
@elvan_hq · X
Feedback surveys with AI summaries for SaaS and ecom teams. NPS, CSAT, CES in minutes, not days. Built for the one-person CX stack who doesn't have time for enterprise tools.
从真实用户获得您的着陆页或SaaS应用的视频评价。
@JessePeplinski · X
building and
对电影和剧集评分、追踪观看进度、建立片单、与朋友对比品味偏好。
@BenjiNaesen · X
I vibe-coded an alternative to Letterboxd last night. You're welcome to try it out. 👀
查看LLM模型在10个基准问题上的评分和排名。
fristovic · HN
She watched me look at model rankings and asked what do the numbers mean... I literally had no good way of explaining it to her so I just came up with something that is approximately in the same ballpark as some of the benchmarks out there lol
检查论文中的AI写作和抄袭，在提交到Turnitin前自查。
PaperCheck — 论文提交前自查工具，提供 Turnitin 查重 + AI 写作检测报告，帮助学生在正式提交前发现问题
在浏览器中运行AI模型基准测试以检测性能回归。
pepperpoppins · HN
Trunchbull, run real models against any benchmark in your browser