Unverified50% confidenceFactExact time
In LJMs, at each event time driven by a Poisson process, the state undergoes a learned instantaneous jump.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn LJMs, the latent state evolves smoothly between events according to closed-form deterministic dynamics.68% similarUnverifiedLatent Jump Models (LJMs) are proposed to address the gap in continuous-time latent state-space models, which cannot directly characterize finite, unobserved shocks.67% similarUnverifiedLSTM 通过引入记忆单元与门控机制（输入门、遗忘门、输出门）来捕捉时间序列中的长期依赖关系64% similarUnverifiedDuring LLM training, memorization occurs, where models may verbatim store fragments of training data and reproduce them during inference.64% similarUnverifiedLLM 在处理复杂任务时容易出现幻觉和逻辑漂移，这与其自回归逐 token 预测机制有关，早期错误会被后续内容放大63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1001485API
curl https://kongchang.com/api/v1/knowledge/claims/1001485MCP
get_claim(id=1001485)