GLM 5.2 (Max reasoning)

4 atoms · 跨 1 天 · 首见 2026-06-25 · 最近 2026-06-25

三色: 🟦 fact 2 · 🟥 take 2 · stance ▲4/▼0/◆0

来源: X 4

时态: fresh:4

标签: 好数字:2 · 好观点:1 · 好思考:1

叙事 (narrative) (1)

• 2026-06-25 · 可靠性评价
  The most reliable agent we've seen
  tags: 好观点 · → daily
  value: direction=最可靠

预测 (forecast) (1)

• 2026-06-25 · 对自动化研究的贡献预期
  I think it'll nontrivially contribute to automated research
  tags: 好思考 · → daily
  value: direction=较大贡献

事实 (fact) (2)

• 2026-06-25 · PostTrainBench 排名
  New #1 on PostTrainBench: GLM 5.2 (Max reasoning) hits 34.29%, narrowly beating Opus 4.8 Max (34.08%)
  tags: 好数字 · → daily
  value: direction=#1
• 2026-06-25 · 测试运行失败率
  zero failed runs across 84 runs (vs ~10% failure rate for Opus agents)
  tags: 好数字 · → daily
  value: qty=0%

时间轴 (近 20)

• 2026-06-25 · fact · PostTrainBench 排名 · → daily
• 2026-06-25 · fact · 测试运行失败率 · → daily
• 2026-06-25 · narrative · 可靠性评价 · → daily
• 2026-06-25 · forecast · 对自动化研究的贡献预期 · → daily

← 实体目录 · 系统日志