English(EN) Even at 8B, LFM2 8B A1B scores 50.5% on MMLU-Pro and only 1% on Long Context Reasoning — a gap that says more about the test than the model. https:// olud.ai/le

LFM2 8B A1B模型基准测试结果喜忧参半

作者 PulseAugur 编辑部 · [1 个来源] · 2026-08-27 01:00

LFM2 8B A1B模型在MMLU-Pro基准测试中取得了50.5%的分数。然而，其在长上下文推理方面的表现却异常低下，仅获得1%的分数。这种差异表明评估方法可能存在问题，而不仅仅是反映了模型的实际能力。 AI

影响这项研究突显了当前大型语言模型（LLM）评估方法可能存在的局限性，尤其是在处理长上下文任务时。

排序理由该条目讨论了一个开源模型的基准测试结果，属于研究范畴。[lever_c_demoted from research: ic=1 ai=1.0]

在 Mastodon — fosstodon.org 阅读 →

模型发布

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

本文如何被排名

Signal score: 16 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该条目讨论了一个开源模型的基准测试结果，属于研究范畴。[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; 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]

Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] · 2026-08-27 01:00

即使是8B模型，LFM2 8B A1B在MMLU-Pro上得分50.5%，在长上下文推理上仅得1%——这一差距更多地说明了测试本身而非模型。https:// olud.ai/le

Even at 8B, LFM2 8B A1B scores 50.5% on MMLU-Pro and only 1% on Long Context Reasoning — a gap that says more about the test than the model. https:// olud.ai/leaderboard.html # LLM # Benchmarks # OpenSource # AI

链接 olud.ai/leaderboard.html

本文如何被排名

报道来源 [1]

即使是8B模型，LFM2 8B A1B在MMLU-Pro上得分50.5%，在长上下文推理上仅得1%——这一差距更多地说明了测试本身而非模型。https:// olud.ai/le

相关实体

相关话题