Curriculum Reinforcement Learning via Morphology-Environment Co-Evolution

no code implementations • 21 Sep 2023 • Shuang Ao, Tianyi Zhou, Guodong Long, Xuan Song, Jing Jiang

Throughout long history, natural species have learned to survive by evolving their physical structures adaptive to the environment changes.

reinforcement-learning Reinforcement Learning (RL)

Empirical Optimal Risk to Quantify Model Trustworthiness for Failure Detection

no code implementations • 6 Aug 2023 • Shuang Ao, Stefan Rueger, Advaith Siddharthan

We propose the Excess Area Under the Optimal RC Curve (E-AUoptRC), with the area in coverage from the optimal point to the full coverage.

Building Safe and Reliable AI systems for Safety Critical Tasks with Vision-Language Processing

no code implementations • 6 Aug 2023 • Shuang Ao

Although AI systems have been applied in various fields and achieved impressive performance, their safety and reliability are still a big concern.

Image Captioning Out of Distribution (OOD) Detection +2

Two Sides of Miscalibration: Identifying Over and Under-Confidence Prediction for Network Calibration

1 code implementation • 6 Aug 2023 • Shuang Ao, Stefan Rueger, Advaith Siddharthan

Then we utilize the class-wise miscalibration score as a proxy to design a calibration technique that can tackle both over and under-confidence.

Confidence-Aware Calibration and Scoring Functions for Curriculum Learning

1 code implementation • 29 Jan 2023 • Shuang Ao, Stefan Rueger, Advaith Siddharthan

In this paper we integrate notions of model confidence and human confidence with label smoothing, respectively \textit{Model Confidence LS} and \textit{Human Confidence LS}, to achieve better model calibration and generalization.

text-classification Text Classification

CO-PILOT: COllaborative Planning and reInforcement Learning On sub-Task curriculum

1 code implementation • NeurIPS 2021 • Shuang Ao, Tianyi Zhou, Guodong Long, Qinghua Lu, Liming Zhu, Jing Jiang

Next, a bottom-up traversal of the tree trains the RL agent from easier sub-tasks with denser rewards on bottom layers to harder ones on top layers and collects its cost on each sub-task train the planner in the next episode.

Continuous Control reinforcement-learning +1

EAT-C: Environment-Adversarial sub-Task Curriculum for Efficient Reinforcement Learning

no code implementations • 29 Sep 2021 • Shuang Ao, Tianyi Zhou, Jing Jiang, Guodong Long, Xuan Song, Chengqi Zhang

They are complementary in acquiring more informative feedback for RL: the planning policy provides dense reward of finishing easier sub-tasks while the environment policy modifies these sub-tasks to be adequately challenging and diverse so the RL agent can quickly adapt to different tasks/environments.

reinforcement-learning Reinforcement Learning (RL)

Learning ULMFiT and Self-Distillation with Calibration for Medical Dialogue System

no code implementations • ICNLSP 2021 • Shuang Ao, Xeno Acharya

In this paper, we investigate the well-calibrated model for ULMFiT and self-distillation (SD) in a medical dialogue system.

Decision Making Knowledge Distillation +2