Expired50% confidenceFactExact time
The core upgrade of Matrix-3.5 is the injection of a memory system into the world model, enabling the model to remember previous scene states in order to maintain coherence.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/23/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/21/2026(expired)
Sources
Related Claims
Unverified上下文图谱为智能体提供了介于临时上下文和模型权重微调之间的中间层记忆机制，可在不改变底层模型的前提下持续积累经验71% similarUnverifiedMemory entries are injected into the context window as part of the system prompt; this is essentially a form of persistent prompt engineering that changes the input the model sees rather than the model's parameters.69% similarUnverified在 Cosmonapse 中记忆通过前置钩子和输出钩子实现，记忆逻辑与推理逻辑解耦，模型调用本身保持不变66% similarUnverifiedMemory机制在大语言模型工程中对应系统提示词工程与外部记忆存储的结合，用于模拟无状态模型的持久化状态66% similarUnverifiedMamba's key innovation is the introduction of a selective mechanism that allows matrices B and C to change dynamically depending on the input content, enabling the model to selectively memorize or forget historical information based on the current input.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/598709API
curl https://kongchang.com/api/v1/knowledge/claims/598709MCP
get_claim(id=598709)