Quantum Calibration AI Agents Need a Measurement Contract

Quantum calibration AI agents need every action tied to a measurable signal, a pass condition, and an escalation path. MIT's case shows the contract.

Administrator Administrator Published on 2026-09-09

The DseWiki Incident: When Read-Only Web Wrote

The DseWiki incident shows why read-only web access must be defined by side effects, not HTTP method names. Here is what the public evidence proves.

Administrator Administrator Published on 2026-09-09

Astra 在 ARC-AGI-3 得分 62.7% 或 99.9%:Harness 本身就是评测对象

Astra 在不同 Harness 下得到 62.7% 和 99.9%。本文拆解分数语义、状态管理差异,以及速度和 token 数据的适用边界。

Administrator Administrator Published on 2026-09-05

Astra Scores 62.7% or 99.9% on ARC-AGI-3: The Harness Is Part of the Benchmark

Astra scored 62.7% or 99.9% on ARC-AGI-3 under different harnesses. Here is what changed, what the numbers mean, and what remains unproven.

Administrator Administrator Published on 2026-09-05

Claude Formalized Fermat's Last Theorem in 11 Days: What the Lean Verification Stack Actually Proves

Claude formalized Fermat's Last Theorem in Lean in 11 days. Here is what the 13-million-line proof and its verification stack establish.

Administrator Administrator Published on 2026-09-05

持久 Agent 需要两份合同:什么时候醒,醒来后能做什么

持久 Agent 需要分别约束唤醒时机与行动权限。一次唤醒只允许观察,任何外部副作用都要重新校验授权。

Administrator Administrator Published on 2026-09-05

Persistent AI Agents Need Separate Wake and Action Contracts

Persistent AI agents need separate contracts for wake timing and action authority. This checklist makes long-running agents safer.

Administrator Administrator Published on 2026-09-05

组织第二大脑如何向专家学习,同时避免固化专家错误

组织第二大脑需要把专家反馈转化为带来源、版本、验证结果和退役规则的知识变更,才能在扩散经验的同时控制错误。

Administrator Administrator Published on 2026-09-03

How to Build an Organizational Second Brain Without Freezing Expert Mistakes

Build an organizational second brain with provenance, versioned knowledge, expert review, replay tests, and safe retirement instead of static RAG.

Administrator Administrator Published on 2026-09-03

EvoUndo:给自我修改的 Agent Harness 加一道可恢复性门禁

EvoUndo 说明能力提升不足以批准 Agent Harness 自我修改。持久变更合并前应通过状态见证、独立影响审计与反事实恢复测试。

Administrator Administrator Published on 2026-09-02
Previous Next