AI 有没有可能直接修改另一个 AI 的神经网络参数，甚至借此让它变得更聪明？11 月 2 日，麻省理工学院（MIT）计算机科学与人工智能实验室（CSAIL）的一项研究在技术圈引发热议。研究团队让 GPT-6 Astra 直接分析并修改另一个 AI ...

本文讲清 Harness² 解决什么问题、机制怎么跑通、实测数据、社区落地与未来边界。 做 Agent 的同学大概都有过这种感受：同一个大模型，换一套提示词、工具编排和记忆策略，干活稳定性就能差出一大截。模型没变，结果先变了。 模型表现拉胯时，很多人的第 ...

Researchers applied an offline fine-tuned language model and association rule mining to extract critical production factor ...

In inference optimization, reward hacking happens when an agent improves its score by exploiting the evaluation without delivering the improvement we actually wanted. This can happen at the kernel ...

System logs generated by globally distributed microservices often hide subtle degradations (like memory thrashing or systemic ...

Salesforce's SRD distills hindsight into foresight, rescuing agent RL when rewards vanish and lifting success from 0% to 60.6 ...

Problem-solving is the process of identifying a problem, generating potential solutions, evaluating them, and implementing ...

On November 2, 2026, four researchers including Rikiya Takehi from MIT CSAIL (Computer Science and Artificial Intelligence ...

Harness engineering is not over. Rather, the "next chapter" has begunThe term "harness engineering" has finally taken root in ...

In the strongest example of "weight space programming", as the technique is called, Astra’s intervention raised Qwen3.5-0.8B’s MATH-500 score from 7.9% to 48.1% while touching approximately 0.016% of ...

OpenCode MCP setup for v2.2.6: build a read-only Python server, lock tools down, pool connections and run headless in CI in ...

Tsinghua's HarnessPro cuts AI coding agent costs 26% while boosting accuracy without model retraining. Factory Droid 1.0 ...