Claims Brief
Microsoft's Temporal GraphRAG Claims Partition Retrieval by Period
The pending claims cover period-specific concept graphs, community summaries and query routing, including incremental updates that avoid rebuilding earlier periods.
Patent Register
24 records indexedClaims Brief · Lead record
What UiPath's Auto-Generated Agent-Tool Application Claims
Claim 1 recites initializing an agent, entering an agentic loop and executing a tool containing both an API definition and an implementation.
AI
Claims Brief
Three independent claims, one architecture, and a title that claims nothing
The published application recites an encoder, a language model and a module between them. What the dependent claims name — and what the title does not limit — is where the scope actually sits.
Claims Brief
Sony Application Claims a Generative AI Method Where Output Quality Tracks the Authentication Key
US20260205268A1, published July 16, 2026, recites embedding an authentication key inside a generative model's own output, with output quality higher when the key is valid than when it is not. It is classified under cryptography, not machine learning.
Claims Brief
Microsoft Application Claims Lookup-Table Method for Mixed-Precision Matrix Multiplication
US20260203368A1, published July 16, decomposes an n-bit weight matrix into one-bit matrices and retrieves precomputed partial products from a lookup table indexed by the weights themselves. Three independent claims, one of them a bit-serial hardware system.
Claims Brief
Just Published: An AI-Agent Permission Claim, Read Through Claim 1
A pending application published July 2, 2026 claims a computing system that resolves an AI agent's natural-language 'semantic entitlement' through a generative model to grant resource access. Here is what the independent claim actually covers — and the fact that it is published, not granted.
Claims Brief
Just Issued: What Claim 1 of NVIDIA's Invertible Speech-Synthesis Patent Actually Covers
A granted patent — issued, not merely published — covers a speech vocoder trained by mapping audio to Gaussian values through invertible layers, then run in reverse to synthesize voice. Here is what the independent claim requires, and where it lands in the G10L/G06N landscape.
Claims Brief
A Newly Published NVIDIA Application Claims Using a Vision Language Model to Spot Image Imperfections
A patent application published June 25, 2026 is directed to prompting a vision language model to evaluate whether an image contains imperfections and to score its quality. It is a pending application, not a granted patent, classified under CPC G06V 10/82.
Claims Brief
A Newly Issued Patent Claims Distilling a Server Speech Model Into On-Device Encoders Without Moving Raw Audio
A patent granted June 23, 2026 and assigned to Google LLC claims a federated distillation method that compresses a large server-side speech recognition encoder into a smaller on-device one using principal component analysis. This is an issued patent, not a pending application, and it lands in a recent Google cluster pointed at on-device, privacy-preserving model training.
Landscape Report
Who Owns the Patents on Transformers and Attention? Reading the G06N Landscape Through Real Grants
A landscape read of who holds attention- and transformer-related neural-network patents, grounded in real granted records across Microsoft, Google, Qualcomm, and others — and a caution about what patent counts do and do not show.
Explainer
Claiming a Neural Network: How 35 U.S.C. 112 Shapes Machine-Learning Patent Claims
Beyond eligibility, AI claims must satisfy 35 U.S.C. 112 — written description, enablement, definiteness, and the means-plus-function rule. An explainer on why machine-learning claims are drafted the way they are.
Explainer
Can an AI Be a Named Inventor on a Patent? What Thaler v. Vidal and USPTO Guidance Establish
Under US law a patent inventor must be a natural person, so an AI system cannot be named — but AI-assisted inventions are not categorically unpatentable. An explainer on Thaler v. Vidal and the USPTO's 2024 inventorship guidance.
Primer
What CPC Class Are AI Patents Filed Under? A Primer on G06N and the Machine-Learning Subgroups
A reference primer on the Cooperative Patent Classification scheme for artificial intelligence: what G06N covers, how its subgroups divide neural networks, learning methods, and machine learning, and where AI spills into G06F and G06V.
Explainer
AI Patent Subject-Matter Eligibility: How the Alice/Mayo Test Decides Whether a Machine-Learning Claim Survives 35 U.S.C. 101
An explainer on why so many machine-learning inventions face a 35 U.S.C. 101 challenge, and how the USPTO's two-step Alice/Mayo framework actually evaluates them under MPEP 2106.
Research Brief
New Off-Policy Evaluation Method Targets Rewards That Go Missing Not at Random
An arXiv preprint proposes an estimator for evaluating offline reinforcement learning policies when logged rewards are systematically missing, and tests it on simulated data and MIMIC-III sepsis records.
Research Brief
Percolation Clusters Proposed as a Structured Testbed for Interpretability Methods
A new arXiv preprint introduces synthetic datasets built on critical mean-field percolation clusters, arguing that today's toy datasets lack the hierarchical, multi-scale structure that real data has.
Claims Brief
What a New Mixture-of-Experts Routing Application Claims, and Where It Lands in G06N
A freshly published application is directed to deciding which experts a downstream transformer block should run, using a perplexity measurement taken mid-network. It is a pending application, not a granted patent.
Research Brief
Decoupling the Task From the Robot: A Meta-RL Framework That Reuses Knowledge Across Bodies
A new arXiv paper learns task-level knowledge on a simplified agent and transfers it to physically different robots through a semantic-magnitude interface, reporting large reductions in tracking error and a fraction of the usual interaction data.
Research Brief
EvolveNav Gives Navigation Agents a Rule Memory That Rewrites Itself at Test Time
A new arXiv paper proposes a zero-shot object-goal navigation framework that mines actionable rules from past trajectories, retrieves them with an upper-confidence-bound strategy, and forecasts outcomes before acting — reporting a 10.1% gain in success rate.
Research Brief
Ternary Mamba Cuts the Token Bill for 1.58-Bit State Space Models by 1,000x
A new arXiv paper shows that ternary state space models can be made from a pretrained checkpoint with quantization-aware training and distillation, instead of training from scratch on 150 billion tokens — and surfaces a new instability called zero-ratio collapse.
Research Brief
A PDE Replaces Score Matching: Kolmogorov Regression Targets the Drift in Diffusion Policies
A new arXiv paper recasts diffusion-policy training as a deterministic boundary-value problem in a Cameron-Martin space, reporting convergence guarantees and a reward-free failure detector, with gains on a manipulation benchmark and a manufacturing-line task.
Research Brief
Fixed-Point Reasoners Use Convergence Itself as a Transformer's Stop Signal
FPRM, a looped transformer described in a new arXiv paper, halts when its latent state reaches a fixed point — letting the model spend more compute on hard reasoning problems and less on easy ones, with results reported on Sudoku, Maze, state-tracking, and ARC-AGI.
Research Brief
Looped World Models Claim 100x Parameter Efficiency by Recycling One Transformer Block
A new arXiv paper proposes LoopWM, a world-model architecture that iterates a single parameter-shared transformer block to refine latent states, treating loop depth as a new scaling axis distinct from model size.
Grant Watch
Autodesk's Just-Issued US 12,657,468 Trains Graph Neural Nets to Read 3D CAD Geometry
A G06N 3/084 grant issued June 16 gives Autodesk a method for training machine learning models — built around a graph neural network — to perform tasks on B-rep CAD objects represented as graphs plus 2D UV-grids.