👉 @ai_python ✍️

لینک مقاله در آرشیو :

https://arxiv.org/html/2509.01187v1


The paper introduces StoxLSTM, a novel stochastic extension of the xLSTM architecture for time series forecasting. By integrating stochastic latent variables into a state space modeling framework, StoxLSTM captures complex temporal patterns and uncertainties more effectively than traditional models. Extensive experiments across diverse datasets show that it consistently outperforms state-of-the-art baselines in both accuracy and robustness.

👉 @ai_python ✍️

پادکست مصنوعی توضیحات به فارسی :

https://t.me/navidcasts/26

ویدیو با زمان کوتاه تر از پادکست صوتی به صورت خلاصه تر و به فارسی :

https://youtu.be/xN6nFUGeXrk?si=aUGPUomtp_6yMFg1