It introduced the process of building AI systems tailored to data and work environments in nuclear, public, and defense sectors.
Both studies achieved an F1 score of 90.14% on the ADReSSo benchmark, surpassing the previous best record
Applying FLUX.2 VAE and Muon improves training efficiency and strengthens understanding of Korean instructions
Naver Cloud proposed four execution principles centered on role division and maintaining control to build sovereign defense AI.
Naver Cloud optimized the Mamba-2 kernel for B200 GPUs using Helion DSL, increasing single-layer computation speed by 1.33x.
NAVER has unveiled 'Sommelier,' a data pipeline designed to minimize latency in real-time voice conversation AI.
Enterprise AI work tools that reduce repetitive tasks and information fragmentation are gaining attention.
Data centers, the core of AI computation, are evolving through cooling, power, and control technologies.
It presents a future in which countless robots are unified and controlled in the cloud through ARC and ARC mind.
The core of AI performance lies in AI semiconductors, which handle fast computation and efficiency.
Hallucination, bias, jailbreaking, and sycophancy in generative AI are threatening safety and trust.
Given only a goal, agentic AI plans, uses tools, and verifies results all on its own.
This proposes an evaluation benchmark for inclusive AI agents that help elderly industrial workers.
Naver has unveiled an omnimodal AI connecting text, images, and audio with 32B Think and 8B Omni.
AI technology, which began with calculators, has evolved through deep learning into omnimodal models that integrate language and the senses.
Team Naver increased multimodal model training throughput by 13.3% by resolving image token imbalance.
Naver Cloud has proposed MERIT, a new LLM post-training strategy that prevents conflicts between datasets and reduces communication costs.
NAVER has developed HyperCLOVA X SEED 4B, a lightweight omni-modal model with enhanced Korean language and multimodal understanding capabilities, equipped with proprietary vision and audio encoders.
Naver Cloud has proposed a technology that detects LLM hallucinations by leveraging future context, without needing internal model information.
Naver Cloud has joined NVIDIA's Nemotron Coalition, beginning collaboration to advance HyperCLOVA X and secure sovereign AI.
At GTC Taipei 2026, Naver Cloud unveiled a full-stack AI partnership with NVIDIA to build a Global AI Factory.
LingoQ turns the English expressions users ask AI about during work into personalized quizzes to help them review.
Separating Encoder, Decoder, and LLM and applying llm-d and USP optimization increased throughput by 2.1x.
Autiverse is an AI tool that turns an autistic teen's day into a comic diary structured around ABCE.
AI Briefing