Latest Articles for Evaluation

Sparse autoencoders enhance the interpretability of AI systems and their decision-making processes.

A look at how AI models grasp essential knowledge of the world.

New benchmark assesses toxicity in large language models across various languages.

This article discusses the need for better evaluation practices in fuzzing research.

This study assesses saliency methods in NLP through human evaluation.

Introducing PQAH for better understanding of AI heatmaps and their evaluation.

A new method enhances optimization in costly high-dimensional problems.

A new method for assessing language models' alignment with human values.

A new method improves image creation from multiple text prompts.

An overview of behaviors in crowdsourcing communities and their impacts.

This research highlights the need for better evaluation of dialogue systems' use of conversation history.

AdvEval exposes weaknesses in Natural Language Generation evaluation metrics.

New tool converts sketches into clear graphics programs for researchers.

A new method enhances trustworthiness of AI outputs in blockchain environments.

Participants tackle the restoration of degraded images in a competitive setting.

A novel system tracks and recognizes dynamic 3D scenes using a single video.

Evaluating algorithms for effective musical phrase segmentation and structure analysis.

A new method improves how intelligence messages are assessed by prioritizing credibility.

New resources enhance assessment of Korean language models.

This article examines a new way to create algorithms with LLMs.

Learn how seven-valued logic enhances decision-making with multiple criteria.

A challenge focusing on deep generative models for realistic medical image generation.

A model assesses the readability of Wikipedia articles across 14 languages.

A new approach using LLMs to create distractors with minimal human input.

A new approach to evaluating biases in automated AI evaluation metrics.

New methods aim to enhance reasoning capabilities in language models.

New metrics shed light on the limitations of language models in representing reality.

A new system for assessing language models using real-world data streams.

Introducing IrokoBench to improve LLM evaluation in African languages.

The ULS23 Challenge aims to improve tumor segmentation in CT scans for better cancer care.

A fresh approach improves detection of fake images created by AI.

A new benchmark aims to assess MLLMs in video understanding across multiple topics.

This study presents a new method for identifying key training images in AI-generated visuals.

Exploring the significance of unlearning methods in modern machine learning.

Examining the key issues in offline MARL and proposing standardized solutions.

Learn about CGP, its function, advantages, applications, and challenges in programming.

A new dataset improves coherence in image-text sequences for effective content creation.

SciEx reveals strengths and challenges of LLMs in scientific evaluation.

SEACrowd aims to improve AI representation for Southeast Asian languages and cultures.

A study evaluates language models on handling multiple tasks simultaneously.