# Few-Shot Adaptation to Non-Stationary Environments via Latent Trend Embedding for Robotics

arXiv: https://arxiv.org/abs/2603.10373 · HTML: https://mlsanity.com/paper/2603.10373/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 29 references that are in the map.
- First posted on arXiv 11 Mar 2026.

## Builds on (most cited first, 22 of 22)

- [Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks](https://mlsanity.com/paper/1703.03400/) (2017)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://mlsanity.com/paper/1704.04861/) (2017)
- [Progressive Neural Networks](https://mlsanity.com/paper/1606.04671/) (2016)
- [Experience Replay for Continual Learning](https://mlsanity.com/paper/1811.11682/) (2018)
- [Learning under Concept Drift: A Review](https://mlsanity.com/paper/2004.05785/) (2020)
- [Conditional Neural Processes](https://mlsanity.com/paper/1807.01613/) (2018)
- [Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables](https://mlsanity.com/paper/1903.08254/) (2019)
- [Foundation Models in Robotics: Applications, Challenges, and the Future](https://mlsanity.com/paper/2312.07843/) (2023)
- [RoboTAP: Tracking Arbitrary Points for Few-Shot Visual Imitation](https://mlsanity.com/paper/2308.15975/) (2023)
- [From Concept Drift to Model Degradation: An Overview on Performance-Aware Drift Detectors](https://mlsanity.com/paper/2203.11070/) (2022)
- [Continual Learning and Catastrophic Forgetting](https://mlsanity.com/paper/2403.05175/) (2024)
- [LOTUS: Continual Imitation Learning for Robot Manipulation Through Unsupervised Skill Discovery](https://mlsanity.com/paper/2311.02058/) (2023)
- [Bridging the Sim-to-Real Gap with Dynamic Compliance Tuning for Industrial Insertion](https://mlsanity.com/paper/2311.07499/) (2023)
- [Continual Domain Randomization](https://mlsanity.com/paper/2403.12193/) (2024)
- [DEAR: Disentangled Environment and Agent Representations for Reinforcement Learning without Reconstruction](https://mlsanity.com/paper/2407.00633/) (2024)
- [GOATS: Goal Sampling Adaptation for Scooping with Curriculum Reinforcement Learning](https://mlsanity.com/paper/2303.05193/) (2023)
- [Few-Shot Learning of Force-Based Motions From Demonstration Through Pre-training of Haptic Representation](https://mlsanity.com/paper/2309.04640/) (2023)
- [Toward Understanding Key Estimation in Learning Robust Humanoid Locomotion](https://mlsanity.com/paper/2403.05868/) (2024)
- [Task and Domain Adaptive Reinforcement Learning for Robot Control](https://mlsanity.com/paper/2404.18713/) (2024)
- [DefGoalNet: Contextual Goal Learning from Demonstrations For Deformable Object Manipulation](https://mlsanity.com/paper/2309.14463/) (2023)
- [Incremental Learning of Full-Pose Via-Point Movement Primitives on Riemannian Manifolds](https://mlsanity.com/paper/2312.08030/) (2023)

## Abstract (arXiv)

Robotic systems operating in real-world environments often suffer from concept shift, where the input-output relationship changes due to latent environmental factors that are not directly observable. Conventional adaptation methods update model parameters, which may cause catastrophic forgetting and incur high computational cost. This paper proposes a latent Trend ID-based framework for few-shot adaptation in non-stationary environments. Instead of modifying model weights, a low-dimensional environmental state, referred to as the Trend ID, is estimated via backpropagation while the model parameters remain fixed. To prevent overfitting caused by per-sample latent variables, we introduce temporal regularization and a state transition model that enforces smooth evolution of the latent space. Experiments on a quantitative food grasping task demonstrate that the learned Trend IDs are distributed across distinct regions of the latent space with temporally consistent trajectories, and that few-shot adaptation to unseen environments is achieved without modifying model parameters. The proposed framework provides a scalable and interpretable solution for robotics applications operating across diverse and evolving environments.