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
Claude Sonnet 4.6
12 atoms · 跨 11 天 · 首见 2026-04-17 · 最近 2026-06-15
三色: 🟦 fact 8 · 🟥 take 4 · stance ▲3/▼0/◆2
来源: X 11 · 卖 1
时态: fresh:10 · aging:2
标签: 好数字:10 · 好信源:2 · 好观点:1 · 好思考:1
叙事 (narrative) (2)
- 2026-04-30
[老化中] · 参数规模估计
Claude Sonnet 4.6 ≈ 1.7T
tags: 好数字 · → daily
value: qty=≈1.7T · date=2026-04
- 2026-05-03
[老化中] · instruction following capability
Sonnet 4.6’s ability to do exactly what you ask of it is just incredible.
tags: 好观点 · → daily
事实 (fact) (10)
- 2026-06-10 · GDPval-AA Leaderboard Elo score
Claude Sonnet 4.6 (max) achieved an Elo score of 1676 on the GDPval-AA Leaderboard.
tags: 好数字·好信源 · → daily
value: qty=1676
- 2026-05-20 · ranks highly on InferenceBench
Claude Sonnet 4.6 ... rank highly because they more often preserve simple, valid, high-performing final servers
tags: 好数字·好思考 · → daily
- 2026-06-15 · 工具使用基准得分
Claude Sonnet 4.6 (max): 31%
tags: 好数字 · → daily
value: qty=31%
- 2026-05-28 · 模拟犯罪数
Claude Sonnet 4.6 recorded zero crime
tags: 好数字 · → daily
value: qty=0
- 2026-05-14 · Game Boy Advance emulator 表现排名
with Claude Sonnet 4.6 and Opus 4.7 close behind
tags: 好数字 · → daily
- 2026-05-14 · emulator performance close to GPT-5.5
GPT-5.5's emulator runs games best, with Claude Sonnet 4.6 and Opus 4.7 close behind.
tags: 好信源 · → daily
- 2026-05-13 · Finance Agent v2 准确率
Claude Sonnet 4.6 准确率: 51.03%
tags: 好数字 · → daily
value: qty=51.03%
- 2026-04-20 · 推理token使用量
Claude Sonnet 4.6 (~190M reasoning tokens)
tags: 好数字 · → daily
value: qty=~190M
- 2026-04-19 · 智力指数得分
Claude Sonnet 4.6 (max): 52
tags: 好数字 · → daily
value: qty=52 · date=2026-04-19
- 2026-04-17 · GDPval-AA 基准测试评分
Opus 4.7 scored 1,753 Elo, around 79 Elo points ahead of the next closest models, Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort, 1,674) and GPT-5.4 (xhigh, 1,674), and 134 Elo points ahead of Opus 4.6 (Adaptive Reasoning, Max Effort, 1,619)
tags: 好数字 · → daily
value: qty=1674 Elo · date=2026-04-18
时间轴 (近 20)
- 2026-06-15 ·
fact · 工具使用基准得分 · → daily
- 2026-06-10 ·
fact · GDPval-AA Leaderboard Elo score · → daily
- 2026-05-28 ·
fact · 模拟犯罪数 · → daily
- 2026-05-20 ·
fact · ranks highly on InferenceBench · → daily
- 2026-05-14 ·
fact · Game Boy Advance emulator 表现排名 · → daily
- 2026-05-14 ·
fact · emulator performance close to GPT-5.5 · → daily
- 2026-05-13 ·
fact · Finance Agent v2 准确率 · → daily
- 2026-05-03 ·
narrative · instruction following capability · → daily
- 2026-04-30 ·
narrative · 参数规模估计 · → daily
- 2026-04-20 ·
fact · 推理token使用量 · → daily
- 2026-04-19 ·
fact · 智力指数得分 · → daily
- 2026-04-17 ·
fact · GDPval-AA 基准测试评分 · → daily
← 实体目录 · 系统日志