InkubaLM aims to improve language processing for underrepresented African languages.
― 7 min read
Cutting edge science explained simply
InkubaLM aims to improve language processing for underrepresented African languages.
― 7 min read
CoRA enhances efficiency in training large language models using shared knowledge.
― 5 min read
TSO enhances language models by focusing on diversity, validity, and adaptability in preference data.
― 7 min read
A study on false refusals in language models and their impact on user experience.
― 6 min read
This project enhances text correction in Bulgarian historical documents using OCR technology.
― 5 min read
LongRecipe improves language models' understanding of long texts efficiently.
― 5 min read
Innovative lightweight transducer enhances speech recognition efficiency and accuracy.
― 6 min read
A look into the effectiveness of pipeline versus end-to-end systems in summarizing across languages.
― 6 min read
A new approach to enhance decoder models for different dialects.
― 5 min read
YA-TA offers personalized support for students and instructors in large classrooms.
― 7 min read
Using AI to simulate data for analyzing emotional connections in mental health.
― 10 min read
Combining SAM and MLLMs for better object localization in images.
― 8 min read
A new model improves depression detection in social media posts with clear explanations.
― 5 min read
This article presents a new method for realistic dialogue systems using user-specific traits.
― 4 min read
A new method improves the reliability of large language models' answers.
― 5 min read
Examining LLMs for generating audio programming code using visual languages.
― 5 min read
This article compares discrete and continuous speech representations for effective speech recognition.
― 5 min read
Introducing HTLA, a model improving text classification accuracy through better label alignment.
― 7 min read
This article discusses a new rating system for evaluating language models more fairly.
― 5 min read
Updates enhance the FLORES dataset for Hausa, Northern Sotho, Xitsonga, and isiZulu.
― 5 min read
A model improving sentence parsing by focusing on entity structure.
― 5 min read
Streamlined systems for improved interactions between humans and machines.
― 5 min read
This study evaluates how well VLMs can understand visual perspectives.
― 5 min read
New methods enhance how tools are selected for language models.
― 8 min read
Examining LVLMs' effectiveness in generating multilingual art explanations.
― 7 min read
A hybrid agent for the Werewolf game enhances interaction and gameplay.
― 6 min read
The study evaluates how well language models handle Hakka cultural understanding.
― 5 min read
Addressing biases in chatbots through regular audits to align with societal values.
― 4 min read
A new method enables language models to correct their own mistakes in math.
― 5 min read
A new benchmark evaluates biases in language models used for medical diagnoses.
― 5 min read
A new design to improve long-term memory in language models.
― 5 min read
ERFSL streamlines reward function creation using large language models.
― 5 min read
Examining the efficiency and latency challenges of SMoE models in language processing.
― 6 min read
A new model improves language processing by focusing on input representation.
― 6 min read
CHESS improves efficiency of language models while maintaining performance on resource-limited devices.
― 6 min read
A new framework combines humor theories with machine learning for effective humor detection.
― 8 min read
New method enhances efficiency of large language models by focusing on relevant information.
― 6 min read
This study investigates the performance of entity linking models in conversational contexts.
― 6 min read
Learn how keyphrase prediction enhances content organization and retrieval.
― 5 min read
New framework enhances sign language recognition through context and visual inputs.
― 6 min read