Latest Articles for Model Evaluation

NPHardEval4V assesses reasoning capabilities of multimodal large language models.

A new dataset to assess planning skills of language models in real-life tasks.

Introducing adversarial hypervolume to better assess deep learning model performance.

This work analyzes the performance of simplified transformers in forecasting tasks.

A new benchmark assesses continual learning in multimodal language models.

A look into PAC-Bayes and its impact on model performance.

AVIBench tests LVLMs to ensure they withstand adversarial visual instructions.

This article reviews the strengths and weaknesses of the VMamba model.

A study comparing multilingual and monolingual models' explanations and their faithfulness.

A new method to assess novelty in generative AI outputs.

Explore various models used for data classification and uncertainty estimation.

A new dataset aims to improve hate speech detection models for the German language.

This paper examines how data affects the evaluation of NLP models.

IsoBench evaluates how models handle text and images to identify strengths.

Learn about adversarial attacks and their impact on machine learning models.

A study comparing the safety performance of popular language models.

A framework to assess how training data influences AI model behavior.

A new benchmark evaluates language models' understanding of word meanings and relationships.

A method for verifying model reliability without true labels.

A study comparing Instance and Neuron Attribution methods in language models.

Exploring how transfer learning impacts model effectiveness across different data contexts.

Introducing the FB method for better model assessment in cosmology.

A study reveals overconfidence issues in AI language and vision models.

This article discusses early stopping to improve model selection efficiency in machine learning.

Exploring the benefits and challenges of shared variable embeddings in machine learning.

New techniques enhance reliability and simplicity in genetic programming models.

Introducing AnyLoss, transforming metrics into loss functions for better model training.

This article discusses new methods for explaining AI decisions in object detection.

A look into how adversarial examples challenge AI models.

Learn key methods for selecting tuning parameters in data analysis for better predictions.

A new benchmark for assessing LLMs in cybersecurity tasks.

This paper proposes new methods to evaluate information fragmentation in machine learning.

This paper introduces an approach for creating easy-to-understand AI classifiers.

This study examines how well pretrained models cluster unseen data.

Introducing new methods to improve forgetting processes in contrastive learning models.

An overview of SVM techniques for handling class imbalance in machine learning.

Tackling the issues of OOD generalization and feature contamination in AI models.

This article explores improvements in sparse autoencoders and their impact on language understanding.

A study on the effectiveness of various lightweight models in image classification.

Introducing a method to evaluate model resilience against data poisoning attacks.