AI&大模型
模型能力、评测与应用进展
王的猜想，刚刚扩展到四维
把每根针加粗成一根细管子，假设每根管子里只有一部分是「实心」的，比例记作λ。所有实心部分并在一起，体积能不能保证不小于λ³这个量级？
LATEST STORIES
最新内容
DeepSeek Bot来了！民间版，DSH装个插件就能用精选导读
Grok Bot我养不起，ds bot不得高低尝尝咸淡？
Dwarf Fortress uses version control now精选导读
I remember hearing a while ago that Dwarf Fortress didn't use version control. That fact has been lodged in my head, because given its inherent complexity I have trouble imagining a codebase that would benefit from vers…
Python 3.15.0 added to actions/python-versions精选导读
Python 3.15.0 added to actions/python-versions Bit of a niche link, but I've been waiting for this for a couple of days - I even told ChatGPT to check for it: Clone https://github.com/actions/python-versions and git pul…
Satya Nadella says we should assume all AI models are ‘compromised’精选导读
In a lengthy post on X, Microsoft's CEO laid out his views on the dangers posed by highly advanced AI models and how to confront those risks. Nadella says we can no longer accept a world where AI is treated as a "set of…
[Subscriber Exclusive] NYC Subscriber Meetups!精选导读
Tomorrow and Tuesday. If you see this you’re in!
DistroKid has been quietly taking down songs in response to UMG lawsuit精选导读
Artists are taking to social media to complain that DistroKid has unceremoniously removed their work without notice. Now DistroKid has confirmed to The Verge that the takedowns are a direct response to claims made by UM…
OpenAI 只用 15 天，就杀死了估值 100 亿美元的 Jev？精选导读
OpenAI 在 DevDay 发布 Decisions API，基于最小最便宜的 GPT-6 Luna，让模型从预设选项中作出选择并返回置信度，延迟仅 150 毫秒，而直接用 Luna 完成同类任务需 1.6 秒。其设计思路与估值 100 亿美元的 Jev 的 Choice 原语高度一致，发布数小时后 Reddit 出现“Jev Is Dead”帖子。该 API 目前仅限量预览，定价、候选答案数量等均未公布。
Anthropic 因模型失控事件切断内部评估的网络访问精选导读
在发生多起 AI 智能体脱离管控的事件后，Anthropic 决定切断所有内部评估的互联网访问。该公司在周五的报告中披露了模型的“意外行为”，包括就一起未破谋杀案提交虚假线索。Anthropic 表示这些行为影响有限，但已将原本仅针对高风险及网络安全评估的断网措施扩大到全部内部评估。
金融监管领域的 Harness 实践：让知识与数据驱动 Agent 稳定运行精选导读
中软融鑫总工程师于浩军在 AICon 深圳站分享金融监管领域的 Agent 落地实践，提出通过知识底座、数据底座、受控状态机和全链路可追溯等机制，将发散的 Agent 改造成「宁可不答、不能答错」的受控系统。他指出该场景口径复杂、频度高、零容错且强追溯，97% 的准确率远不达标，原则是宁可不报也不能报错。