Author: Dan Petrovic

  • Revealed: The exact search result data sent to Google’s AI.

    Revealed: The exact search result data sent to Google’s AI.

    —

    by

    Dan Petrovic
    in ,

    UPDATE: Addressing guardrails, hallucinations and context size. 1. People are reporting difficulties in recreating the output due to guardrails and hallucinations. 2. Snippet context sometimes grows to several chunks. Guardrails Google attempts (and in many cases) succeeds at blocking these requests, but it does so in a very clumsy way so that we actually get…

  • Beyond Rank Tracking: Analyzing Brand Perceptions Through Language Model Association Networks

    Beyond Rank Tracking: Analyzing Brand Perceptions Through Language Model Association Networks

    —

    by

    Dan Petrovic
    in , , ,

    This post is based on the codebase and specifications for AI Rank, an AI visibility and rank tracking framework developed by DEJAN AI team: https://airank.dejan.ai/ Abstract: Traditional SEO has long relied on rank tracking as a primary metric of online visibility. However, modern search engines, increasingly driven by large language models (LLMs), are evolving beyond…

  • Teaching AI Models to Be Better Search Engines: A New Approach to Training Data

    Teaching AI Models to Be Better Search Engines: A New Approach to Training Data

    —

    by

    Dan Petrovic
    in

    A recent patent application* reveals an innovative method for training AI models to become more effective at understanding and answering human queries. The approach tackles a fundamental challenge in modern search technology: how to teach AI systems to truly understand what people are looking for, rather than just matching keywords. The Core Innovation The traditional…

  • Self-Supervised Quantized Representation for KG-LLM Integration

    Self-Supervised Quantized Representation for KG-LLM Integration

    —

    by

    Dan Petrovic
    in

    Paper: https://arxiv.org/pdf/2501.18119 This paper proposes a method called Self-Supervised Quantized Representation (SSQR) for seamlessly integrating Knowledge Graphs (KGs) with Large Language Models (LLMs). The key idea is to compress the structural and semantic information of entities in KGs into discrete codes (like tokens in natural language) that can be directly input into LLMs. Here’s a…

  • What does Gemini think about your brand?

    —

    by

    Dan Petrovic
    in

    Inside Chrome Dev, there’s a quantized version of Google’s flagship model Gemini for those who have it enabled. The model does many things from summarization, translation, writing assistance all the way to scam prevention. The model definition is a secret, but its weights are stored as a 3GB .bin file on the user machine. Inside…

  • Google’s Privacy Sandbox: Navigating the Cookieless Future

    Google’s Privacy Sandbox: Navigating the Cookieless Future

    —

    by

    Dan Petrovic
    in

    The digital advertising landscape is undergoing a significant transformation as privacy concerns grow and regulations like GDPR and CCPA take effect. Third-party cookies, long the backbone of online advertising, are being phased out due to their intrusiveness and potential for misuse. In response, Google has introduced the Privacy Sandbox, a collection of initiatives aimed at…

  • Why deep learning works.

    —

    by

    Dan Petrovic
    in

    Here’s a powerful excerpt from “Deep Learning with Python” by François Chollet”: The nature of generalisation in deep learning has rather little to do with the deep learning models themselves and much to do with the structure of the information in the real world. The input to an MNIST classifier (before preprocessing) is a 28 × 28 array of…

  • Introducing VecZip: Embedding Compression Algorithm

    Introducing VecZip: Embedding Compression Algorithm

    —

    by

    Dan Petrovic
    in , ,

    Embeddings are vital for representing complex data in machine learning, enabling models to perform tasks such as natural language understanding and image recognition. However, these embeddings can be massive in size, creating challenges for storage, processing, and transmission. At DEJAN AI, we’ve developed VecZip, a novel approach to address this issue, and reduce the file size…

  • Site Engagement Metrics

    Site Engagement Metrics

    —

    by

    Dan Petrovic
    in

    To access the feature in Chrome visit: chrome://site-engagement/ Google Site Engagement Metrics Framework plays a crucial role in assessing and analyzing user engagement with websites. This framework leverages detailed metrics, such as user interactions and engagement scores, to provide insights into browsing behavior. Here’s a breakdown of how this system works, based on the Site…

  • Beyond Links: Understanding Page Transitions in Chrome

    Beyond Links: Understanding Page Transitions in Chrome

    —

    by

    Dan Petrovic
    in

    When SEOs think about user behavior, the conversation often revolves around clicks, links, and conversions. But in Chrome, there’s an underlying layer of data that tells a much richer story—page transitions. These are the bread and butter of how users navigate, revealing not just where they go, but how they got there. For SEOs, understanding…