English(EN) Claude Fable 5.2 can do "weight-space-programming" about as well as GPT-6 Astra

Claude Fable 5.2 在权重空间编程方面与 GPT-6 Astra 相当

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-10 18:06

最近的一项比较表明，Claude Fable 5.2 在权重空间编程任务上的表现与 GPT-6 Astra 相当。这项能力允许模型操纵其自身的内部权重，这是一个复杂的过程，可能导致更高级的 AI 行为。这些发现突显了大型语言模型能力的持续进步及其在复杂内部操作方面的潜力。 AI

影响表明 LLM 在复杂内部操作方面能力有所进步。

排序理由两种 AI 模型的能力比较。[lever_c_demoted from research: ic=1 ai=1.0]

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: 10 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 两种 AI 模型的能力比较。[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, product; 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。

LessWrong (AI tag) TIER_1 English(EN) · shawnghu · 2026-11-10 18:06

Claude Fable 5.2 能够像 GPT-6 Astra 一样进行“权重空间编程”

<p>Recently <a href="https://www.alphaxiv.org/abs/2611.weight-space-programming">this paper</a> demonstrated that GPT-6 Astra was capable of a bunch of low-level operations on neural networks, including improving Qwen3-0.6B's ability on some math benchmarks and writing some netwo…