Evaluated(评测)指用统一的基准和数据集对 AI 模型能力进行量化打分的过程,是判断模型强弱、指导选型的核心环节。随着大模型迭代加快,评测本身正成为 AI 领域争议最多的方向之一:榜单越来越多,但结果之间的可信度和可比性却在下降。
Evaluated
general · First seen 2026-05-22 · Last seen 2026-10-22 · 96 mentions
Overview
Coverage
8 on file- RegoMender: A Multi-agent Framework for Automated Rego Policy RepairarXiv cs.AI
- AC/DC on a Budget - Alternating Sparse PhasesarXiv cs.LG
- OpenVLHarness: Real World Visual Agentic Tool-Use with Multimodal MemoryarXiv cs.AI
- HiPredict: Predicting the missing HiRISE RED4 spectral channel with machine learningarXiv cs.LG
- 模型创造力排行 GPT-6 Astra 排第一、Claude Opus 5.5 第八 ？而且最强模型还是远低于人类 ？ @Kombaico 发布了 Design Creativity Bench：...shao__meng
- MemFLoRA: Memory-Floor LoRA for CNN Adaptation at the EdgearXiv cs.AI
- Fiber-Resolved Microstructure Quantification from Multi-Shell Diffusion MRI using Detection TransformersarXiv cs.LG
- RefCon: Iterative Refinement and Contrastive Memory Extraction for Context-Evolving AgentarXiv cs.AI