non-stationarity

Tag

Cards List
#non-stationarity

In-Context Reinforcement Learning under Non-Stationarity: A Survey

arXiv cs.AI · 2026-07-15 Cached

This survey examines in-context reinforcement learning (ICRL) under non-stationary environments, where a pretrained decision model adapts through accumulated context without parameter updates. It organizes the literature around what changes, how it unfolds, and how observable it is, and identifies research gaps such as stale-context stress tests and adaptive forgetting.

0 favorites 0 likes
#non-stationarity

StateFlow: Dual-State Recurrent Modeling for Long-Horizon Time Series Forecasting

arXiv cs.LG · 2026-07-02 Cached

This paper introduces StateFlow, a recurrent forecasting framework that extends the Variability-Aware Recursive Neural Network (VARNN) to long-horizon multivariate time series forecasting by using a dual-state recurrent backbone and a chunk-based decoder, achieving competitive performance against strong baselines.

0 favorites 0 likes
#non-stationarity

Semantics-Enhanced Retrieval-Augmented Time Series Forecasting

arXiv cs.AI · 2026-06-16 Cached

Proposes SERAF, a multimodal retrieval-augmented framework for time series forecasting that uses both numerical similarity and self-generated textual descriptions to retrieve historical patterns, improving forecasting under non-stationarity. Experiments on seven real-world datasets show effectiveness over state-of-the-art baselines.

0 favorites 0 likes
#non-stationarity

Stationarity-Aware Retrieval-Augmented Time Series Forecasting

arXiv cs.LG · 2026-06-04 Cached

SARAF is a Stationarity-Aware Retrieval-Augmented Forecasting framework that adaptively balances relevance and diversity in retrieval for time series forecasting, modulating diversification strength based on dataset-level stationarity to handle non-stationary regime shifts. Accepted to KDD 2026, it demonstrates competitive performance over strong baselines on eight real-world datasets.

0 favorites 0 likes
#non-stationarity

Position: Deployed Reinforcement Learning should be Continual

arXiv cs.LG · 2026-06-04 Cached

This position paper argues that deployed RL agents should never stop learning, as the train-then-fix paradigm inherently fails to address non-stationarity and distribution shift in real-world environments. The authors identify four sources of post-deployment non-stationarity and advocate for continual RL as the standard approach for deployed systems.

0 favorites 0 likes
#non-stationarity

Solipsistic Superintelligence is Unlikely to be Cooperative

arXiv cs.AI · 2026-06-03 Cached

This paper argues that superintelligent AI systems designed under a solipsistic paradigm that treats the world as stationary will be self-undermining and uncooperative, leading to collective failures. The authors call for a new research paradigm that treats interdependence and cooperation as core design principles.

0 favorites 0 likes
← Back to home

Submit Feedback