Kimi 推出的新模型 Kimi K2 拥有 1 万亿参数，并在 HLE 基准测试中取得了最先进（SOTA）的成果。它还在 BrowseComp 和 TauBench 中展示了能力。另外，Soumith Chintala 已从 PyTorch 离职。 AI
影响 在 HLE 基准测试上设定了新的 SOTA，可能影响未来的模型开发。
排序理由 新模型发布，参数量和基准测试性能显著。
AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →
Kimi K2 模型拥有 1T 参数和 SOTA HLE，同时 Soumith Chintala 离开 PyTorch
本文如何被排名
- Signal score
- 0 / 100
- Composite score across the factors below. Higher = stronger signal that this story matters right now.
- Newsworthiness bucket
- Frontier Release
- 新模型发布，参数量和基准测试性能显著。
- 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, paper
- 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
- 336 days old
- Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.
完整方法见我们的编辑标准。
报道来源 [1]
-
Kimi K2 Thinking：1T-A32B参数，SOTA HLE，BrowseComp，TauBench && Soumith 离开 Pytorch
**Moonshot AI** launched **Kimi K2 Thinking**, a **1 trillion parameter** mixture-of-experts (MoE) model with **32 billion active experts**, a **256K context window**, and native **INT4 quantization-aware training**. It achieves state-of-the-art results on benchmarks like **HLE (…