Language and culture internalization for human-like autotelic AI

C Colas, T Karch, C Moulin-Frier… - Nature Machine …, 2022 - nature.com
Building autonomous agents able to grow open-ended repertoires of skills across their lives
is a fundamental goal of artificial intelligence (AI). A promising developmental approach …

Autotelic agents with intrinsically motivated goal-conditioned reinforcement learning: a short survey

C Colas, T Karch, O Sigaud, PY Oudeyer - Journal of Artificial Intelligence …, 2022 - jair.org
Building autonomous machines that can explore open-ended environments, discover
possible interactions and build repertoires of skills is a general objective of artificial …

Safety verification for neural networks based on set-boundary analysis

Z Liang, D Ren, W Liu, J Wang, W Yang… - … Symposium on Theoretical …, 2023 - Springer
Neural networks (NNs) are increasingly applied in safety-critical systems such as
autonomous vehicles. However, they are fragile and are often ill-behaved. Consequently …

BIRDNN: Behavior-Imitation Based Repair for Deep Neural Networks

Z Liang, T Wu, C Zhao, W Liu, B Xue, W Yang, J Wang… - Neural Networks, 2025 - Elsevier
The increasing utilization of deep neural networks (DNNs) in safety-critical systems has
raised concerns about their potential to exhibit undesirable behaviors. Consequently, DNN …

Repairing deep neural networks based on behavior imitation

Z Liang, T Wu, C Zhao, W Liu, B Xue, W Yang… - arxiv preprint arxiv …, 2023 - arxiv.org
The increasing use of deep neural networks (DNNs) in safety-critical systems has raised
concerns about their potential for exhibiting ill-behaviors. While DNN verification and testing …

The ecology of open-ended skill acquisition

C Moulin-Frier - 2022 - inria.hal.science
An intriguing feature of the human species is our ability to continuously invent new problems
and to proactively acquiring new skills in order to solve them: what is called open-ended skill …

Two-Stream Heterogeneous Graph Network with Dynamic Interactive Learning for Video Question Answering

M Peng, X Shao, Y Shi, X Zhou - 2023 International Joint …, 2023 - ieeexplore.ieee.org
Video question answering (VideoQA) challenges the joint learning of visual and linguistic
knowledge. Whilst the dynamic video-question interaction is not well explored in previous …

Verifying safety of neural networks from topological perspectives

Z Liang, D Ren, B Xue, J Wang, W Yang… - Science of Computer …, 2024 - Elsevier
Neural networks (NNs) are increasingly applied in safety-critical systems such as
autonomous vehicles. However, they are fragile and are often ill-behaved. Consequently …

Credit assignment for trained neural networks based on Koopman operator theory

Z Liang, C Zhao, W Liu, B Xue, W Yang… - arxiv preprint arxiv …, 2022 - arxiv.org
Credit assignment problem of neural networks refers to evaluating the credit of each network
component to the final outputs. For an untrained neural network, approaches to tackling it …

UR4NNV: Neural Network Verification, Under-approximation Reachability Works!

Z Liang, T Wu, R Zhao, B Xue, J Wang, W Yang… - arxiv preprint arxiv …, 2024 - arxiv.org
Recently, formal verification of deep neural networks (DNNs) has garnered considerable
attention, and over-approximation based methods have become popular due to their …