LLM 可直接读写神经网络权重，RSI 能力实现从 0 到 1 突破
原标题：[AINews] LLMs can read & write neural network weights: "Weight Space Programming" cuts Autoresearch out of the RSI loop
AI 摘要
Latent Space 报道了一篇关于「权重空间编程」的论文：研究者让大语言模型直接读取并编辑较弱模型的神经网络权重，无需编写 PyTorch/JAX 代码或运行训练流程，即可识别权重功能、根据失败轨迹一次性修复缺陷，甚至从零生成完整网络权重并超过部分独立训练的模型。该能力被视为递归自我改进（RSI）从 0 到 1 的突破，已有独立复现，但目前仅适用于简单游戏和强化学习策略，尚难扩展到更复杂任务。同期 DeepSeek 开源了「超级智能体」harness 原型，并出现 V4.1 命名混淆。
正文节选
Amid a deepening bear market for AI/tech (worthy of dark humor) with Anthropic bonds receiving a “junk” rating and OpenAI fundraise headlines, we continue to see very surprising capability finds in the community, such as today. The best way to put today’s blockbuster paper in context is to recall Karpathy’s autoresearch from a mere 8 months ago, which presaged a summer of recursive self improvement fears: Abstractly, the demonstration is that you can take a sufficiently smart model and through r