PulseAugur
中
实时 03:28:30
English(EN) 📊 K2-V2 (Medium) scores are out, measured independently, not from its own team. GPQA 59.8%, MMLU-Pro 76.1%, but Humanity's Last Exam drops to 4.4% — the gap bet

K2-V2 (Medium) 基准测试分数揭示推理差距

作者 PulseAugur 编辑部 · [1 个来源] ·

对 K2-V2 (Medium) 模型的独立评估揭示了其在各种基准测试上的表现。虽然它在 GPQA 上获得了 59.8% 的分数，在 MMLU-Pro 上获得了 76.1% 的分数，但其在 Humanity's Last Exam 上的表现下降至 4.4%。这表明该模型自我报告的能力与其真实的推理表现之间存在显著差异。 AI

影响 凸显了大型语言模型（LLM）在推理能力方面可能被高估的现象，表明需要更严格、独立的评估。

排序理由 AI 模型的独立基准测试结果。[lever_c_demoted from research: ic=1 ai=1.0]

在 Mastodon — mastodon.social 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

K2-V2 (Medium) 基准测试分数揭示推理差距

本文如何被排名

Signal score
11 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
AI 模型的独立基准测试结果。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, other
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

  1. Mastodon — mastodon.social TIER_1 English(EN) · opensourceaitech ·

    📊 K2-V2 (Medium) 分数已公布，由第三方独立测量，非其团队内部数据。GPQA 59.8%，MMLU-Pro 76.1%，但 Humanity's Last Exam 下降至 4.4% — 差距

    📊 K2-V2 (Medium) scores are out, measured independently, not from its own team. GPQA 59.8%, MMLU-Pro 76.1%, but Humanity's Last Exam drops to 4.4% — the gap between self-reported and real-world reasoning is where the story is. See the full breakdown. https:// olud.ai/leaderboard.…

    链接 olud.ai/leaderboard.html

相关实体

相关话题