Graham Neubig

A new method to detect and classify errors in language model outputs.

TroVE simplifies function creation for programming tasks, improving accuracy and efficiency.

New model improves interlinear glossing for endangered languages.

Analyzing RAG systems to improve language model performance in question answering.

A new method enhances communication skills of language agents.

Discover how tools enhance language model capabilities and performance.

A new method supports the preservation of at-risk languages through detailed documentation.

An overview of different learning methods used by language models.

DataTune improves synthetic dataset generation for better language model training.

A new method improves code generation accuracy using external documents.

Learn how inference-time algorithms enhance text generation performance.

A method to create tailored datasets for answering questions using diverse information.

A new method for generating videos using both text and images.

A new method allows language models to generate their own training data for better performance.

This system translates English speech to German text instantly for seamless communication.

A new method enhances agents' abilities to complete complex digital tasks efficiently.

MORCELA adjusts language model scores to better reflect human language judgment.

Evaluating language models' abilities in synthetic data creation using AgoraBench.

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

Discover how software engineering agents are transforming coding efficiency.