Tech blog
Show category:
-
All
{
AI
AI Agent
Artificial Intelligence
Business insights
Category 1
Category 2
Category 3
Classification
Classifier
Fraud
Generative AI
Geometric deep learning
Large language models
Life Science
Machine learning
Market segmentation
Named Entity Recognition
Natural language processing
NLP
Object detection
Recommendation System
Retrieval Augmented Generation
Segmentation
Sentiment Analysis
Speech Recognition
Supply Chain
Synthetic data
Topic Modeling
User Engagement
User segmentation
Machine Unlearning: Erasing knowledge from LLMs
According to the Open Worldwide Application Security Project (OWASP) report on top 10 LLM risks, the Top-1 risk is prompt injection. But guess what comes second?
Let’s dive into machine unlearning — how can we try to erase knowledge from LLM weights, and how the bad guys might undo all your efforts.
Read moreBuilding a Deep Research Multi-Agent System
In a world overwhelmed by information, our deep research system offers a new approach to document analysis—not by summarising, but by uncovering what’s missing. It identifies knowledge gaps, autonomously formulates questions, and investigates using external and internal sources. Powered by a ReAct-enabled agent, the system mimics a human researcher’s reasoning loop—asking, acting, and adapting to surface insights that otherwise remain hidden.
Enhancing retrieval systems with Domain Adaptation
Given a large, unlabeled, domain-specific set of documents. What is the most effective set of techniques to achieve good retrieval performance?
Retrieval Augmented Generation (RAG) is currently being used across various industries and enterprises as it allows for up-to-date, traceable, and fact-based Large Language Model (LLM) generated answers, crucial in professional settings.
Read more
AI-Generated imagery in digital and print media for Bonnier News
Is it feasible for current image generation models to produce high-quality, photorealistic visual content suitable for both print in glossy magazines, and digital publishing?
In this blog post, we discuss the framework we used to answer this question and provide a Google Colab notebook with Python code for an automated analysis of gender bias in image-generating models. Spoiler alert: an online A/B test carried out by Bonnier News (a controlled experiment where two variations of an ad are shown to different groups of website visitors), revealed a notable preference for AI-generated images. Specifically, ads with AI-generated images showed a markedly higher click-through rate.
Read more