Verification of neural networks: a challenge to overcome

The formal verification of neural networks presents many challenges. Although there are languages to describe how a neural network is expected to behave, current validation tools do not consider the full richness of these languages. CEA-List is investigating practices from the field of programming languages to expand the scope of what can be formally verified.

Uncertainty in AI-guided Monte Carlo simulations

We developed a method, PEM, to adapt the Monte Carlo algorithm and penalize regions that are uncertain for the AI, mitigating AI errors and making deep-learning-based materials modeling trustworthy.

A machine-learning-based methodology for fast and efficient modeling of power consumption in digital architectures

In this research, we proposed an AI-assisted methodology to significantly speed up power consumption modeling. Clustering techniques are used to select representative windows, automatically reducing register transfer level (RTL) traces.

AI algorithms for gamma-ray spectrometry used in field measurement

Incremental software verification for cybersecurity

In cybersecurity, newly-introduced software vulnerabilities must be detected immediately. The generic incremental analysis techniques recently integrated into Frama-C can verify code quickly after local modifications.

Fast fully-homomorphic encryption

In fields like defense, industry, healthcare, and finance, data privacy is a major challenge. One promising avenue is using fullyhomomorphic encryption (FHE) to compute on encrypted data. However, to implement this approach, many hurdles must be overcome—ones that CEA-List has been working on for years.

Hardware accelerator tightly coupled to a RISC-V core for post-quantum cryptography

With the development of quantum computers—and their threat to modern cryptography—post-quantum cryptography is gradually making inroads into the industrial ecosystem. HQC (Hamming Quasi-Cyclic) is one of the post-quantum encryption algorithms recently standardized by NIST.

Estimation of a multifingered gripper joint configuration using 3D point clouds and AI

In robotic gripping, the positions of the effectors (phalanges or fingertips) on an object must be controlled precisely. CEA-List looked at a practical case where the only information available was a 3D point cloud of the gripper, generated by visual sensors, simulations, or generative neural networks.

A cognitive programming interface

CEA-List developed an ontology-based cognitive interface to enable context-sensitive programming that requires no robotics expertise.

Active lubrication, a kinesthetic and vibrotactile haptic brake

CEA-List has been researching the principle of active lubrication for around ten years and has developed a fully integrated actuator that paves the way for a new type of brake suitable for a wide range of applications.