2

2025 Scientific Report • April 1, 2026

AI for faster multi-physics simulation: an architecture for nonlinear structural mechanics

LEBNN, developed by CEA-List, is a neural network architecture for the mechanics of statically hyperelastic structures.

Read more
2025 Scientific Report • April 1, 2026

Embedded non-volatile memory with robust, optimized accesses

The CEA has been conducting R&D on resistive memory (ReRAM)— an alternative to flash memory—for more than a decade. A complete system integrating ReRAM into a processor architecture was recently designed and validated for advanced memory company Weebit Nano.

Read more
2025 Scientific Report • April 1, 2026

Quantum computing applications and use cases

Much like classical application-specific processors, quantum processing units (QPUs) are expected to be used to speed up certain computational steps in algorithm execution.

Read more
2025 Scientific Report • April 1, 2026

Qbricks: toward the formal verification of hybrid classical/ quantum programs

Advances in quantum computing are creating new challenges to our current code validation practices. CEA-List is developing new formal-analysis-based verification techniques to respond to this new context.

Read more
2025 Scientific Report • April 1, 2026

MiRAG: multi-level retrievalaugmented generation for visual question answering

CEA-List developed the MiRAG model for visual question answering about named entities. This is the first time a retrieval augmented generation (RAG)-based approach to generative AI has been applied to this task.

Read more
2025 Scientific Report • April 1, 2026

Annotation-free world discovery: With xMOD, 2D and 3D vision work together

xMOD combines 2D vision (cameras) and 3D vision (LiDAR sensors) in a novel cross-distillation method. The AI learns to segment its environment from motion cues in the images, delivering beyond state-of-the-art performance.

Read more
2025 Scientific Report • April 1, 2026

3D scene analysis using natural language queries

The DiSCO-3D semantic segmentation method is used to discover, in a 3D scene, the elements corresponding to the semantic subconcepts of a user query expressed in natural language.

Read more
2025 Scientific Report • April 1, 2026

Low-latency motion detection with event graphs

Event-driven cameras provide low-latency motion detection. Our method leverages asynchronous event graphs that take full advantage of the cameras’ high time resolution to detect motion with very low latency (just 50 milliseconds) while reducing the number of operations 48-fold compared to the state of the art.

Read more