Tsinghua University's Institute for Artificial Intelligence (AIR) and its collaborators have introduced WAM, a new embodied intelligence framework built around In-Context causal learning. The announcement came via Techub News, with the original report credited to Qbitai. According to the report, WAM delivers significant capability improvements while keeping model parameters frozen. Rather than updating weights, the framework uses causal intervention and dynamic environment adjustment to boost performance. This points to a context-driven mechanism: the model learns from in-context causal relationships and adapts as the environment shifts. The brief notice did not cover additional details such as benchmark results, open-source plans, or specific application scenarios. The work originates from Tsinghua AIR and partner institutions, though those collaborators were not named in the source material.

Beijing-based Zhongshu Ruizhi Technology has secured a strategic financing round worth several hundred million yuan, with backing from China Internet Investment Fund, Suchuangtou National Social Security Fund, Financial Street Capital, ICBC Capital, and Kunlun Capital, while existing shareholders also added to their positions. The deal comes only three months after the company completed a B round worth over 100 million yuan. The company, founded in April 2020 by Han Han, a Tsinghua University PhD and former researcher involved in China’s national big data and AI policy design, focuses on high-reliability decision AI for industrial use cases. Its core pitch centers on self-developed causal AI technologies, including a meta-causal cognitive theory, causal models, and a dynamic ontology engine, which it says address hallucinations, weak reasoning, missing temporal logic, unverifiable decisions, and conflicts among multiple rule sets in generative AI systems. According to the report, Zhongshu Ruizhi has already served more than 50 central state-owned enterprises, state-owned enterprises, and industrial group clients across power, petroleum, aerospace, and other sectors, with deployments in more than 800 highly complex production scenarios. The company said this financing will be used for foundational research, wider rollout of proven applications, overseas expansion, and senior talent recruitment.