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
Fugu
6 atoms · 跨 1 天 · 首见 2026-06-22 · 最近 2026-06-22
三色: 🟦 fact 0 · 🟥 take 6 · stance ▲1/▼4/◆1
来源: X 6
时态: fresh:6
标签: 好观点:4 · 好数字:2
叙事 (narrative) (3)
- 2026-06-22 · cost transparency missing
they literally NEVER REPORT the number of output tokens or cost to achieve a benchmark/task
tags: 好观点 · → daily
- 2026-06-22 · model pool transparency missing
they are super vague and unclear about which model is in the LLM pool (they only mention closed source api one)
tags: 好观点 · → daily
- 2026-06-22 · not a sovereign/open-source AI release
nothing here to celebrate here in regards to sovereign / opensource ai
tags: 好观点 · → daily
预测 (forecast) (1)
- 2026-06-22 · export controls immunity claim
it's technically true that it will never be subject to export controls though!
tags: 好观点 · → daily
事实 (fact) (2)
- 2026-06-22 · Fugu (non-ultra) SWE Bench Pro performance vs Opus
fugu (not the ultra version) is basically a classifier that selects which model at each turn is most likely to answer correctly (in other words a router). this leads to -10 points on SWE Bench pro compared to opus
tags: 好数字 · → daily
value: qty=-10 points · direction=below Opus
- 2026-06-22 · Fugu (non-ultra) performance on other benchmarks vs Opus
gets some gains on other benchmarks but very slight
tags: 好数字 · → daily
value: direction=very slight gains
时间轴 (近 20)
- 2026-06-22 ·
fact · Fugu (non-ultra) SWE Bench Pro performance vs Opus · → daily
- 2026-06-22 ·
fact · Fugu (non-ultra) performance on other benchmarks vs Opus · → daily
- 2026-06-22 ·
narrative · cost transparency missing · → daily
- 2026-06-22 ·
narrative · model pool transparency missing · → daily
- 2026-06-22 ·
narrative · not a sovereign/open-source AI release · → daily
- 2026-06-22 ·
forecast · export controls immunity claim · → daily
← 实体目录 · 系统日志