Naver’s image-translation patent: keeping translated text aligned with a moving camera
A practical reading of US12197883B2: selecting a usable frame, recognising and translating text, tracking its position and refreshing stale results.
Read articleNEWS & INSIGHTS
Ideas, developments
and perspectives on IP.
Analysis of IP practice, AI technology and international protection, alongside news from IPLEX. Explore the articles in English.
360 articles and updates
AllAI & SoftwareDesignsTrademarksPatentsOverseas TrademarksIPLEX NewsOverseas DesignsOverseas PatentsCopyrightTechnology Listing
70 articles
FlashAttention-2: making long-context attention run efficiently on GPUs
A practical reading of Tri Dao’s paper: tiled attention, reduced non-matrix work and better GPU work partitioning, with kernel benchmarks distinguished from full-model training.
Read articleMAD-RAG: preserving visual evidence when retrieved text distracts a model
A technical reading of dual-question inputs and attention mixing, separating benchmark gains, recovered errors and deployment limitations.
Read articleBaidu’s RAG approach to conflicts between model answers and retrieved knowledge
US20260087055A1 distinguishes factual and semantic conflicts and describes different routes to a final answer.
Read articleNVIDIA’s draft-token selection: faster LLM inference by choosing what to verify
An analysis of US20260236763A1: speculative decoding, marginal benefit and processing cost, with a hypothetical delivery-support example.
Read articleDEEPX DX-M1 and edge AI: connecting training with deployment feedback
An analysis of published US application US20260065050A1, using a hypothetical factory inspection workflow to explain heterogeneous processors, model optimisation and evaluation feedback.
Read articleAttention Residuals: How AI combines earlier layers, and what matters for patents
An accessible explanation of Attention Residuals, Block AttnRes, the limits of the reported results and the implementation details relevant to AI patent drafting.
Read articleFiX: fine-grained forgetting in softmax attention
Yongduck Kim examines FiX’s feature-wise gates, numerical implementation and paged cache, distinguishing reported gains from unresolved limitations.
Read articleMHAR: Reading earlier layers through different feature subspaces
Yongduck Kim examines Multi-Head Attention Residuals: depth routing, reported training results, implementation costs and the relationship between technical features and effects.
Read articleColumn: Claude Computer Use and the Data That Trains AI Agents
Writing for AI Times, IPLEX Managing Partner Yongduck Kim examines the training data behind computer-operating AI agents through U.S. Patent No. 12,585,862.
Read articleIP Strategy Begins with Understanding Technology: An IPLEX Interview
Explore how IPLEX combines AI and software expertise with a practical understanding of business strategy. The interview discusses education for founders, mentoring and communication with clients.
Read articleAI Patents: What Can Be Protected and How to Prepare
This guide explains the issues to consider and the materials to prepare when seeking patent protection for AI technology.
Read article