16 stories tagged Multilingual
Researchers propose a meta-learning framework for RLHF and DPO that transfers preference data across languages, enabling effective alignment of large language models in low-resource languages with minimal data.
Researchers created a new test to evaluate how well AI models handle mixed-language text, like Hindi or Tamil blended with English. This is common in many multilingual communities but often confusing for AI systems.
Researchers created a new test showing AI vision systems struggle with languages written in multiple scripts. The study highlights how current AI models unfairly disadvantage billions of people who use different writing systems for the same language. However, the original source does not support testing this with tools like Google Lens or Microsoft Seeing AI, as those are not explicitly mentioned.
PaddlePaddle released PP-OCRv6, an AI that can read text from images in 50 languages. It's up to 700% more efficient than its predecessor, working even on smartphones. This makes it easier to digitize books, signs, and documents in any language.
A new study reveals that current AI models favor languages like English and French, making them less effective and more expensive for speakers of underrepresented languages. Researchers propose new methods to make these models more fair and efficient for everyone.
Researchers introduced AdaMame, a new AI training approach that helps large language models reason better in multiple languages. This could make math-solving AI tools more reliable for non-English speakers.
Researchers created a multilingual dataset to improve AI's ability to share facts across languages. This could make AI assistants more reliable worldwide.
Researchers created IdiomX, a large-scale test for AI to understand and translate idioms. This could help AI communicate more naturally in different languages. Idioms are tricky because their meanings aren't always literal.
Researchers developed DetectRL-X, a benchmark to test AI text detectors in real-world, multilingual scenarios. This tool evaluates detectors across 8 languages, aiming to improve reliability and governance of AI-generated content.
IBM has open-sourced Granite Embedding Multilingual R2, a powerful AI model that improves search and translation across languages. It handles up to 32,000 words at once, making it useful for large documents and complex queries.
Researchers have created DocAtlas, a new framework that can understand documents in over 80 languages, including those with limited resources. This tool could make digital services more accessible to non-English speakers worldwide.
Researchers propose a data-efficient method to train reasoning models to seamlessly switch between languages. This approach could revolutionize multilingual AI applications by leveraging code-switching as a strength rather than an error.
Researchers used Brain Score to evaluate language models trained on diverse languages and structured sequences, finding shared processing properties. The study suggests neural models capture universal linguistic features beyond specific language structures.
NVIDIA has open-sourced a new OCR model that supports multiple languages and leverages synthetic data for training. The model is designed for speed and accuracy in text recognition tasks.
Researchers introduce Claim2Vec, a multilingual embedding model designed to group similar fact-checking claims. This innovation aims to improve automated fact-checking by efficiently clustering recurrent misinformation claims across languages.
Google Translate's Live Translate with Headphones feature is now available on iOS, expanding to more countries. This feature allows real-time translation through headphones, enhancing communication for travelers and multilingual users.