Latest Articles for Benchmark

High energy physics researchers are optimizing software for diverse computing resources.

This approach simplifies choosing effective pretraining datasets for language models.

A new approach to assess AI benchmarks for cultural understanding.

New method generates complete simulations in code from natural language inputs.

This article assesses how well LLMs generate test cases for Java programs.

Research reveals weaknesses in online toxicity detection using ASCII art techniques.

Exploring the performance gap of general models in finance tasks.

Discover the latest improvements in Arabic language processing technology and its impact.

Learn how technology helps edit tiny details in images effectively.

A new benchmark tests AI agents in realistic CRM tasks.

Data contamination impacts the performance of language models and evaluation methods.

This article discusses the need for transparency in language model benchmarks.

Machines learn to connect sound and visuals in 3D spaces.

Transforming complex benchmark data into clear visual insights.

Milabench provides tailored benchmarks to improve AI performance evaluations.

Researchers create tools to improve AI's grasp of the Ukrainian language.

Are NLI tasks still relevant for testing large language models?

Researchers develop a new benchmark for studying low-frequency somatic mutations in genetics.

A look into causal inference methods and the role of Structural Causal Models.

A look into the challenges of matching servers with requests amid uncertainty.

VidHal benchmarks video models' ability to accurately interpret content.

A look into the behavior of Marshak waves under complex conditions.

This article explores improvements in offline reinforcement learning by breaking down actions.

A new method improves counting in images using LVLMs.

Learn how investors can make better payoff choices.

A study on creating efficient document database queries from examples.

A new benchmark reveals gaps in AI 3D spatial reasoning skills.

Researchers adapt language models to improve Dutch fluency, showcasing new techniques.

A new benchmark aims to enhance AI's understanding of scientific charts.

Discover how new methods improve question answering systems for better user experience.

Discover how machines are improving their understanding of images and texts.

How AI models can fake their intelligence through manipulation.

A new framework boosts language models for low-resource languages.

CG-Bench helps machines analyze long videos better with clue-based questions.

A new benchmark to test LLM reasoning across cultural backgrounds.

Examining the capabilities and limitations of AI agents in task automation.

A guide to understanding and addressing faults in deep learning models.

Combining visual data and language models enhances fixing software issues.

Explore how new benchmarks are transforming document interpretation by AI models.