Mrinmaya Sachan

This study examines how language models mirror children's biases in solving math problems.

AFaCTA aids fact-checkers in identifying true and false claims efficiently.

This article examines how language models balance factual and counterfactual information.

This article examines methods to enhance confidence levels in language models.

Combining causal relationships with prediction improves sentiment analysis accuracy.

Examining fairness and ethics of how language models respond to users.

Error Span Annotation offers a fast and reliable approach to translation quality assessment.

DIRAS improves relevance annotation for information retrieval, optimizing performance across various domains.

This article examines how certain neurons affect uncertainty in language model predictions.

Analyzing how LLMs make moral choices across languages and cultures.

Improving tutoring responses with effective error verification techniques for better student support.

Learn how questions enhance reading and comprehension.

Researchers explore using natural language for better model alignment.

Learn key grammar rules to improve your writing and communication.

Learn how the sequence of information affects AI answer quality.