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
UK AISI
7 atoms · 跨 5 天 · 首见 2026-04-30 · 最近 2026-07-02
三色: 🟦 fact 6 · 🟥 take 1 · stance ▲2/▼0/◆1
来源: X 7
时态: fresh:7
标签: 好信源:7 · 好数字:2 · 好观点:1
叙事 (narrative) (1)
- 2026-06-27 · 被用户称赞为goated
_@gbrl_dick UK AISI is so goated_
tags: 好观点·好信源 · → daily
事实 (fact) (6)
- 2026-07-02 ⭐ · 发现计算投入规模影响性能
spending 1M vs 100M matters
tags: 好数字·好信源 · → daily
- 2026-07-02 ⭐ · 发现上下文长度影响性能
having 1M vs 100K context matters
tags: 好数字·好信源 · → daily
- 2026-06-02 · Mythos 사이버보안 성능 평가
영국 정부 산하 AI 안전연구소(UK AISI)는 미소스가 복잡한 사이버보안 업무에서 오퍼스보다 훨씬 뛰어난 성능을 보인다고 평가했다.
tags: 好信源 · → daily
value: compare=Claude Opus
- 2026-05-13 · 此前测试的模型版本
UK AISI previously tested a partially trained version. The latest results are on the actual Mythos Preview -- the model as it was on the day we launched Glasswing (on April 7th).
tags: 好信源 · → daily
- 2026-06-27 · benchmark结果
the UK’s governmental assessment shows the same thing
tags: 好信源 · → daily
- 2026-04-30 · cyber range test results
This is just one eval, but it's an important one - UK AISI’s cyber range tests long-horizon, agentic capability. 5.5 performs similarly to Mythos.
tags: 好信源 · → daily
value: qualifier=5.5 performs similarly to Mythos
时间轴 (近 20)
- 2026-07-02 ·
fact · 发现计算投入规模影响性能 · → daily
- 2026-07-02 ·
fact · 发现上下文长度影响性能 · → daily
- 2026-06-27 ·
narrative · 被用户称赞为goated · → daily
- 2026-06-27 ·
fact · benchmark结果 · → daily
- 2026-06-02 ·
fact · Mythos 사이버보안 성능 평가 · → daily
- 2026-05-13 ·
fact · 此前测试的模型版本 · → daily
- 2026-04-30 ·
fact · cyber range test results · → daily
← 实体目录 · 系统日志