Get the newsletter
Topic · AI / ML

Latest AI patents, Page 201

How the biggest tech companies are building AI: language models, assistants and agents, vision, voice, and the systems that train them. The artificial intelligence and machine learning patent applications we have broken down from the companies we track, newest first.

in the Nov 12 release
96
in November so far
188
filings tracked
2,637
sub-topics
10

Recent filings

Newest first · 2,637 total
  1. IBM · Published May 21

    IBM Patents an LLM Loop That Fixes Its Own Failing API Tests

    Writing test cases for APIs is tedious — and when they fail, debugging them is even worse. IBM wants to hand that entire feedback loop to an LLM and let it fix its own mistakes.

  2. IBM · Published May 21

    IBM Patents a Deduplicated Mixture-of-Experts System for Multi-Model AI Inference

    Running multiple AI models simultaneously means loading a lot of redundant components into expensive GPU memory — IBM's patent describes a way to store overlapping parts just once and share them across every model that needs them.

  3. IBM · Published May 21

    IBM Patents a Parallel Attention Model for Multi-Channel Time Series Analysis

    Most transformer models look at data one way at a time. IBM's new patent describes a system that simultaneously asks 'when did this happen,' 'which sensor said it,' and 'what does the content mean' — all at once, in parallel.

  4. IBM · Published May 21

    IBM Patents a Self-Tuning System for LLM RAG Pipeline Parameters

    Getting an LLM to answer questions accurately over your own documents depends heavily on a tangle of configuration knobs — chunk size, overlap, retrieval depth — and IBM is patenting a system that uses machine learning to find the best combination automatically, without a human having to guess.

  5. Adobe · Published May 21

    Adobe Patents a Context-Aware Query Rewriter That Routes Requests to the Right AI Model

    When you're mid-conversation with an AI assistant and type something vague like 'make it shorter,' the system has to figure out what 'it' means — and then decide which AI model is even equipped to handle the job. Adobe's new patent tries to solve both problems at once.

  6. Adobe · Published May 21

    Adobe Patents a Neural Material Encoding System for Lightweight 3D Assets

    Adobe wants to let you take a richly detailed 3D object — complex curves, intricate surface materials — and compress its visual fidelity into a lightweight package that runs comfortably on a phone or in a browser tab.

  7. Adobe · Published May 21

    Adobe Patents a Parallel Approach to Causal Linear Attention in Neural Networks

    Attention mechanisms are the engine inside modern AI — but they're also the bottleneck. Adobe's new patent describes a way to run a key part of that engine in parallel, potentially making AI inference meaningfully faster without sacrificing accuracy.

  8. Adobe · Published May 21

    Adobe Patents an AI System That Merges Image Content with a Separate Style Source

    Adobe is patenting a way to let an AI take the subject matter from one photo and the visual style from another — and merge them into a single coherent image, without you having to write a prompt describing either.

  9. Adobe · Published May 21

    Adobe Patents a Distillation Trick to Speed Up AI Image Generation

    Diffusion models make beautiful images, but they're slow — they have to reverse a noisy process dozens of times per generation. Adobe's new patent tries to shortcut that bottleneck by teaching a separate encoder to predict where the diffusion process is headed, without actually running all the steps.

  10. Adobe · Published May 21

    Adobe Patents a Condition-Map System for Controlling AI Image Structure

    Adobe is patenting a way to give AI image generators a spatial blueprint — a 'condition map' — so the model produces images that match a specific layout or structure, not just a text description. It's the difference between asking for 'a cityscape' and handing the model a rough sketch of exactly where the buildings should go.

  11. Adobe · Published May 21

    Adobe Patents a Parallel-Batching Fix for LLM Context Window Overflow

    Every LLM has a context window — a hard cap on how much text it can read at once. Adobe's new patent attacks that limit head-on by splitting your reference examples into parallel batches and letting the model process them simultaneously.

  12. Adobe · Published May 21

    Adobe Patents a Fix for AI Images That Mix Up Objects and Locations

    Ask an AI image generator to put 'a red ball on the left and a blue cube on the right' and you'll often get the ball on the right, the cube on the left, or both crammed into the same corner. Adobe's new patent targets exactly that failure mode.