Sep 2026· Zenodo (CERN European Organization for Nuclear Research)
Advanced Memory and Neural Computing
Abstract
The massive growth in computer chip complexity, along with physical limits like heat, has severely strained traditional chip design methods. Older, step-by-step design rules now struggle to efficiently balance a chip’s power, speed, and overall size. Because of this, regular computers (like standard CPUs and GPUs) waste too much time and energy to run fast, non-stop Artificial Intelligence (AI) tasks in the real world. This paper explores “neuromorphic” (brain-like) computing as the ultimate solution for real-time AI. By acting like a human brain and only processing data when a specific event actually happens, these new chips react almost instantly and save huge amounts of battery power. However, building these highly complex circuits is very difficult. To overcome these design bottlenecks, engineers now use advanced AI-driven methods to replace rigid human rules with smart, data-driven optimization models. This paper summarizes how brain-like chips work, compares their benefits to regular computers, and explains the AI design tools needed to successfully build the future of real-time AI hardware.
This paper investigates developing real-world perception for active agents, proposes Gibson Environment for this purpose, and showcases a set of perceptual tasks learned therein.
F. Xia, Amir Zamir, Zhi-Yang He et al.· 2018 IEEE/CVF Conference on...· 1,069 citations· ⚡112
GAOKAO-Bench is introduced, an intuitive benchmark that employs questions from the Chinese GAOKAO examination as test samples, including both subjective and objective questions that contribute a robust evaluation benchmark for future large language models and offers valuable insights into the advantages and limitations...
Xiaotian Zhang, Chun-yan Li, Yi Zong et al.· arXiv.org· 216 citations· ⚡17
An algorithm is proposed which leverages disentangled semantic factors to generate adversarial perturbation by altering controlled semantic attributes to fool the learner towards various "adversarial" targets.
Haonan Qiu, Chaowei Xiao, Lei Yang et al.· arXiv.org· 208 citations· ⚡17
This work proposes a HALO that directly maximizes the utility of generations instead of maximizing the log-likelihood of preferences, and suggests that the best loss depends on the inductive biases most appropriate for a given setting, an oft-overlooked consideration.
This work investigates the possibilities of using LLMs in a resume screening setting via a document retrieval framework that simulates job candidate selection and finds that the MTEs are biased, significantly favoring White-associated names in 85% of cases and female-associated names in only 11.1% of cases.
This work shows that orders of magnitude enhancement in performance could be obtained by a combination of hardware improvements and tight quantum-HPC integration and introduces high-performance architectures for quantum-probabilistic computing with custom-designed accelerators to tackle today's industry-scale classical...
Masoud Mohseni, Artur Scherer, K. Johnson et al.· arXiv.org· 121 citations· ⚡9
With $2.1 million funding from Google.org, the open-source Public Transit Intelligence Hub will unify public transit monitoring, operations, and passenger communication.
