以下部分引用 AI 总结，现阶段 AI 仍然有幻觉。请以内容的原文为准。

数据新鲜度群聊 07-14 ✓卖方 07-14 ✓Wrap 断档15天 ✕News 07-15 ✓X 断档12天 ✕页面生成 07-15

MiniCPM-V 4.6 1.3B Instruct

8 atoms · 跨 1 天 · 首见 2026-05-11 · 最近 2026-05-11

三色: 🟦 fact 6 · 🟥 take 2 · stance ▲3/▼1/◆0

来源: X 8

时态: fresh:7 · aging:1

标签: 好数字:7 · 好观点:2 · 好思考:1

2026-05-11 [老化中] · Pareto frontier extension
MiniCPM-V 4.6 1.3B Instruct extends the open weights Pareto frontier on both Intelligence vs. Total Parameters and Intelligence vs. Active Parameters at the sub-2B scale
tags: 好思考 · → daily
value: qty=sub-2B

2026-05-11 · parameter count
At 1.3B parameters, MiniCPM-V 4.6 1.3B Instruct scores 13 on the Artificial Analysis Intelligence Index, the highest for any open weights model under 2B parameters
tags: 好数字·好观点 · → daily
value: qty=1.3B
2026-05-11 · output token efficiency
MiniCPM-V 4.6 1.3B Instruct is highly token efficient, and used just 5.4M output tokens to run the Artificial Analysis Intelligence Index, the lowest output token count measured for any open weights model under 4.5B scoring 10 or above on the Intelligence Index
tags: 好数字·好观点 · → daily
value: qty=5.4M
2026-05-11 · Intelligence Index score
MiniCPM-V 4.6 1.3B Instruct scores 13 on the Artificial Analysis Intelligence Index
tags: 好数字 · → daily
value: qty=13 · date=2026-05-11
2026-05-11 · model type
MiniCPM-V 4.6 1.3B Instruct is a vision-language model that supports text, image, and video input with text output
tags: 好数字 · → daily
value: qty=1.3B · date=2026-05-11
2026-05-11 · MMMU-Pro score
MiniCPM-V 4.6 1.3B Instruct scores 38% on MMMU-Pro
tags: 好数字 · → daily
value: qty=38%
2026-05-11 · context window size
Context window: 262K
tags: 好数字 · → daily
value: qty=262K
2026-05-11 · knowledge recall score
MiniCPM-V 4.6 1.3B Instruct shows weaknesses on knowledge recall and hallucination avoidance. It scores -85 on AA-Omniscience
tags: 好数字 · → daily
value: qty=-85