Blog

The importance of humans: AI expectations for 2025 and key NeurIPS learnings

Explore the key takeaways from NeurIPS 2024 and how they align with our predictions for the AI landscape in 2025.

Code Runner: Secure, scalable code execution for model evaluation

Meet Code Runner, the new in-platform code execution engine designed to simplify coding-related tasks and deliver higher-quality datasets for coding-related projects.

Bringing AI to the browser: SAM2 for interactive image segmentation

Learn how to leverage the power of the SAM2 in a web browser, enabling interactive image segmentation without the need for powerful servers or specialized hardware.

Advance LLM reasoning with advanced fact-checking and prompt rating tools

Quantumworks Lab's new fact-checking and prompt-rating tools improve LLM accuracy and reasoning capabilities by allowing users to evaluate responses, correct errors, and flag bad prompts.

Leaderboards: Multimodal reasoning now available & updated evaluations for image, speech and video

Learn about the latest updates using expert human evaluations and a scientific approach to rank and measure preference across leading multimodal reasoning, image, speech, and video models. 

Inside the matrix: A look into the math behind AI

Matrices are crucial in AI for processing multi-dimensional data in areas like machine learning and computer vision. They represent linear maps and transform input into output, making them central to many AI methods.

Multi-step reasoning: Teach LLMs to think critically

Train LLMs to perform complex reasoning tasks and provide high quality step-by-step instructions with new "message step tasks" feature in the multimodal chat editor.

Enhanced video editor adds deeply nested classifications to improve visibility and efficiency

Streamline video data labeling with new features for deeply nested classifications and overall improved video editor functionality.

Announcing Quantumworks Lab Alignerr Connect: Discover and hire proven AI experts

Explore Alignerr Connect to find and recruit AI trainers to join your data factory directly and push the boundaries of the AI frontier.

Product spotlight: Delivering high-quality data at scale for GenAI models

Check out our Q3 recap of what's new from Labelbox. We examine updates that improve GenAI model evaluation, deliver high-quality data at scale, and simplify human evaluations.