Unverified50% confidenceFactExact time
The recurrent state space model (RSSM) consists of a deterministic recurrent path (GRU) and a stochastic path (latent variables): the deterministic state h_t is updated by the previous moment's h_{t-1} and action a_{t-1} through an RNN, and the stochastic state z_t is sampled from a prior distribution parameterized by h_t; this hybrid design balances long-range memory and multimodal uncertainty expression.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedDreamerV3采用的RSSM（循环状态空间模型）由Danijar Hafner等人提出，将隐状态分解为确定性循环状态和随机隐变量74% similarUnverifiedDreamer系列（Hafner等人，2019-2023）核心是循环状态空间模型（RSSM），通过重构观测学习表征，属于生成式方法67% similarUnverifiedSSM 通过随时间演化的隐藏状态 h_t = A·h_{t-1} + B·x_t 压缩历史信息，输出由 h_t 线性映射得到，单步解码计算量与序列长度无关，显存占用恒定。66% similarUnverifiedState Space Models (SSMs) pull the direction back toward fixed-size recurrent memory, but use a different state structure and update rules.65% similarUnverifiedState space models (SSMs) have O(1) memory footprint and computation during inference, whereas the KV cache of Transformers grows linearly with sequence length.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1011061API
curl https://kongchang.com/api/v1/knowledge/claims/1011061MCP
get_claim(id=1011061)