Our Machine Learning team has been at the forefront of machine learning developments, including deep learning, support vector machines, and semantic analysis, for over a decade. We develop innovative technologies integrated into NEC’s products and services. Machine learning is the critical technology for data analytics and artificial intelligence. Recent progress in this field opens opportunities for various new applications.
Deep learning will maintain prominence with more robust model architectures, training methods, and optimization techniques. Enhanced interpretability and explainability will be imperative, especially for AI systems in critical domains like healthcare and finance. Addressing bias and ensuring fairness in AI algorithms will be a top priority, leading to the development of tools and guidelines for ethical AI. Federated learning, quantum computing’s potential impact, and the growth of edge computing will diversify ML applications.
Natural language processing will continue to advance, driving progress in conversational AI, while healthcare, finance, education, and creative industries will witness profound AI integration. As quantum computing matures, it could revolutionize machine learning, while edge computing and federated learning will expand AI’s reach across various domains. Our machine learning research will produce innovation across industries, including more accurate medical diagnoses, safer autonomous systems, and efficient energy use while enabling personalized education and AI-generated creativity.
Read our news and publications from our world-class team of researchers from our Machine Learning department.
As 2025 comes to a close, NEC Laboratories America reflects on a year defined by scientific breakthroughs, global collaboration, and real-world impact. Our researchers advanced the state of the art across AI, optical networking and sensing, system security, and multimodal analytics, while expanding our intellectual property portfolio and presence at the world’s leading conferences.
How many mistakes do published AI papers contain? Peer-reviewed publications form the foundation upon which new research and knowledge are built. Errors that persist in the literature can propagate unnoticed, creating confusion in follow-up studies and complicating reproducibility. The accelerating pace of research and the increasing demands on the peer-review system make such mistakes harder to detect and avoid. To address this, we developed a Paper Correctness Checker based on GPT-5 to systematically identify mistakes in papers previously published at top AI conferences and journals. Our analysis focuses on objective mistakes-e.g., errors in formulas, derivations, calculations, figures, and tables-that have a clearly verifiable ground truth. We intentionally exclude subjective considerations such as novelty, importance, or writing quality. We find that published papers contain a non-negligible number of objective mistakes and that the average number of mistakes per paper has increased over time-from 3.8 in NeurIPS 2021 to 5.9 in NeurIPS 2025 (55.3% increase); from 4.1 in ICLR 2018 to 5.2 in ICLR 2025; and from 5.0 in TMLR 2022/23 to 5.5 in TMLR 2025. Human experts reviewed 316 potential mistakes identified by the AI Checker and confirmed that 263 were actual mistakes, corresponding to a precision of 83.2%. While most identified issues are relatively minor, correcting them would reduce confusion in the literature and strengthen reproducibility. The AI Checker also surfaced potentially more substantive mistakes that could affect the interpretation of results. Moreover, we show that the AI Checker can propose correct fixes for 75.8% of the identified mistakes. Overall, this study highlights the potential of frontier LLMs to detect and correct objective mistakes in published papers, helping to establish a firmer foundation of knowledge.
We study the problem of length generalization (LG) in transformers: the ability of a model trained on shorter sequences to maintain performance when evaluated on much longer, previously unseen inputs. Prior work by Huang et al. (2025) established that transformers eventually achieve length generalization once the training sequence length exceeds some finite threshold, but left open the question of how large it must be. In this work, we provide the first quantitative bounds on the required training length for length generalization to occur. Motivated by previous empirical and theoretical work, we analyze LG in several distinct problem settings: error control vs. average error control over an input distribution, infinite-precision softmax attention vs. finite-precision attention (which reduces to an argmax) in the transformer, and one- vs. two-layer transformers. In all scenarios, we prove that LG occurs when the internal behavior of the transformer on longer sequences can be “simulated” by its behavior on shorter sequences seen during training. Our bounds give qualitative estimates for the length of training data required for a transformer to generalize, and we verify these insights empirically. These results sharpen our theoretical understanding of the mechanisms underlying extrapolation in transformers, and formalize the intuition that richer training data is required for generalization on more complex tasks.
Eric Blow of NEC Labs will address how machine-learning methods applied to distributed acoustic-sensing data can monitor facility perimeters and detect intrusion via walk, dig, or drive events over buried optical fibre—for example achieving ~90% classification accuracy.
Large Language Models (LLMs) have demonstrated remarkable capabilities in understanding and generating human-like text, yet they largely operate as reactive agents, responding only when directly prompted. This passivity creates an “awareness gap,” limiting their potential as truly collaborative partners in dynamic human discussions. We introduce , a framework designed to bridge this gap by training models to proactively decide not just to say, but critically, to speak. Our primary contribution is a scalable two-stage data generation pipeline that synthesizes a large-scale dataset of realistic multi-turn human discussions. Each discussion is annotated with one of five intervention types (e.g., Factual Correction, Concept Definition) and contains an explicit conversational trigger where an AI intervention adds value. By training models to predict a special silent token when no intervention is needed, they learn to remain quiet until a helpful contribution can be made. We explore two architectural baselines: an integrated end-to-end model and a decoupled classifier-generator system optimized for low-latency inference. We evaluate these models on their ability to accurately time interventions and generate helpful responses, paving the way for more situationally aware and proactive conversational AI.
Cell fate decisions are highly coordinated processes governed bycomplex interactions among numerous regulatory genes, whiledisruptions in these mechanisms can lead to developmental abnormalitiesand disease. Traditional methods often fail to capture suchcombinatorial interactions, limiting their ability to fully model cellfate dynamics. Here, we introduce MetaVelo, a global feature explanationframework for identifying key regulatory gene sets influencingcell fate transitions. MetaVelo models these transitions as ablack-box function and employs a differentiable neural ordinary differentialequation (ODE) surrogate to enable efficient optimization.By reparameterizing the problem as a controllable data generationprocess, MetaVelo overcomes the challenges posed by the nondifferentiablenature of cell fate dynamics. Benchmarking acrossdiverse stand-alone and longitudinal single-cell RNA-seq datasetsand three black-box cell fate models demonstrates its superiorityover 12 baseline methods in predicting developmental trajectoriesand identifying combinatorial regulatory gene sets. MetaVelo furtherdistinguishes independent from synergistic regulatory genes,offering novel insights into the gene interactions governing cellfate. With the growing availability of high-resolution single-celldata, MetaVelo provides a scalable and effective framework fo
Question Answering (QA) accounts for a significantportion of LLM usage “in the wild”.However, LLMs sometimes produce false ormisleading responses, also known as hallucinations.Therefore, grounding the generatedanswers in contextually provided information—i.e., providing evidence for the generated text—is paramount for LLMs’ trustworthiness. Providingthis information is the task of context attribution.In this paper, we systematically studyLLM-based approaches for this task, namelywe investigate (i) zero-shot inference, (ii) LLMensembling, and (iii) fine-tuning of small LMson synthetic data generated by larger LLMs.Our key contribution is SYNQA: a novel generativestrategy for synthesizing context attributiondata. Given selected context sentences, anLLM generates QA pairs that are supported bythese sentences. This leverages LLMs’ naturalstrengths in text generation while ensuring clearattribution paths in the synthetic training data.We show that the attribution data synthesizedvia SYNQA is highly effective for fine-tuningsmall LMs for context attribution in differentQA tasks and domains. Finally, with a userstudy, we validate the usefulness of small, efficientLMs (fine-tuned on synthetic data fromSYNQA) in context attribution for QA.
Adapting large Video-Language Models (VLMs) for action detection using only a few examples poses challenges like overfitting and the granularity mismatch between scene-level pre-training and required person-centric understanding. We propose an efficient adaptation strategy combining parameter-efficient tuning (LoRA) with a novel learnable internal feature augmentation. Applied within the frozen VLM backbone using FiLM, these augmentations generate diverse feature variations directly relevant to the task. Additionally, we introduce a group-weighted loss function that dynamically modulates the training contribution of each augmented sample based on its prediction divergence relative to the group average. This promotes robust learning by prioritizing informative yet reasonable augmentations. We demonstrate our method’s effectiveness on complex multi-label, multi-person action detection datasets (AVA, MOMA), achieving strong mAP performance and showcasing significant data efficiency for adapting VLMs from limited examples.
We provide quantitative bounds on the length of sequences required to be observed during training for a transformer to length generalize, e.g., to continue to perform well on sequences unseen during training. Our results improve on Huang et al. [8], who show that there is a finite training length beyond which length generalization is guaranteed, but for which they do not provide quantitative bounds.
Designing protein-binding proteins with high affinity is critical in biomedical research and biotechnology. Despite recent advancements targeting specific proteins, the ability to create high-affinity binders for arbitrary protein targets on demand, without extensive rounds of wet-lab testing,remains a significant challenge. Here, we introduce PPDiff, a diffusion model to jointly design the sequence and structure of binders for arbitrary protein targets in a non-autoregressive manner. PPDiff builds upon our developed Sequence Structure Interleaving Network with Causal attention layers (SSINC), which integrates interleaved self-attention layers to capture global amino acid correlations, k-nearest neighbor (kNN) equivariant graph layers to model local interactions in three-dimensional (3D) space, and causal attention layers to simplify the intricate interdependencies within the protein sequence. To assess PPDiff, we curate PPBench, a general protein complex dataset comprising 706,360 complexes from the Protein Data Bank (PDB). The model is pretrained on PPBench and finetuned on two real-world applications: target-protein mini-binder complex design and antigen-antibody complex design. PPDiff consistently surpasses baseline methods, achieving success rates of 50.00%, 23.16%, and 16.89% for the pretraining task and the two downstream applications, respectively.
