In the Weights is your new AI-centric vanity search

2026-06-20 · Source: AI News & Artificial Intelligence | TechCrunch · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Fundamental Awareness, short

Summary

The "In the Weights" website, created by Thomas Dimson and Joey Flynn, offers a novel "vanity search" experience by measuring how well AI models recall individuals without web search tools. The platform queries various models, including Grok, Gemini, multiple GPT versions, Claude, and Llama, with questions like "Who is X?" It then clusters descriptions and assigns a "strength score" indicating an individual's prominence within AI model weights. For instance, a tech blogger received a score of 641, placing them in the top 6%, while Macaulay Culkin and Luciano Pavarotti scored 988. The site also highlights model-specific responses and potential hallucinations, such as GPT-5.4 Mini's ambiguous description for "Anthony Ha." Dimson stated the project emerged from a belief that Google vanity searches are outdated in 2026 as more traffic shifts to LLMs, aiming to explore how lives are encoded in AI's numerical parameters. Future plans include investigating model biases and discrepancies.

Key takeaway

For AI product managers evaluating public perception or data scientists assessing model biases, "In the Weights" offers a unique lens into how individuals are represented within large language models. You should consider using such tools to understand the implicit knowledge encoded in your models, beyond explicit training data. This can reveal unexpected recall patterns, potential biases, or hallucination tendencies, informing your model evaluation and ethical AI development strategies.

Key insights

"In the Weights" quantifies an individual's recall prominence within AI model parameters, reflecting a shift from traditional web search.

Principles

• AI model weights encode human existence.
• LLM traffic shifts redefine digital prominence.
• Model recall varies, showing biases and hallucinations.

Method

The platform queries multiple LLMs with identity questions, clusters similar descriptions, and assigns a numerical "strength score" based on recall and confidence.

In practice

• Check your "strength score" across various LLMs.
• Compare model recall for different public figures.
• Identify specific model hallucination patterns.

Topics

• AI Model Weights
• Large Language Models
• Digital Identity
• AI Hallucinations
• Model Bias
• Vanity Search

Best for: Entrepreneur, Tech Journalist, General Interest, AI Product Manager

Related on AIssential

• Website "In the Weights" shows whether AI models know who you are — The site reveals how deeply individuals are embedded in LLM training data, quantified by a "strength score".
• Cognitive Effects and Biases in Large Language Models — This tutorial clarifies cognitive effects and biases in LLMs by bridging psychology and NLP.
• How Open-Weight Models Changed the AI Landscape — Open-weight models drive rapid AI innovation through shared designs and iterative improvements.
• From Exams to Escape Rooms: How We Learned to Test AI — AI evaluation is a dynamic process, constantly adapting to measure increasingly complex model capabilities beyond narrow task performance.
• Contribution Weights: A Geometrical Analysis of Self-Attention Transformers — Contribution Weights offer a geometrically-informed, more faithful measure of token importance and reveal active roles for attention sinks in LLMs.
• From Feelings to Metrics: Understanding and Formalizing How Users Vibe-Test LLMs — Formalized "vibe-testing" bridges the gap between LLM benchmark scores and real-world user utility by personalizing evaluation.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI News & Artificial Intelligence | TechCrunch.