A new method blends weak and strong AI models to align with human values.
― 8 min read
Computer Science - Computation and Language
RSSAMGPT provides precise support for researchers in metal additive manufacturing.
― 5 min read
SCALM improves caching in chat services to enhance efficiency and reduce costs.
― 5 min read
Exploring tensor attention and its impact on data processing in AI models.
― 4 min read
A new method enhances the alignment of language models using multiple references.
― 7 min read
This research focuses on generating pseudo-programs to enhance reasoning tasks in models.
― 5 min read
This study evaluates ASR systems' performance with individuals who stutter.
― 7 min read
This article examines how attacks affect LLM safety and response generation.
― 5 min read
A universal audio clip can mute advanced ASR models like Whisper.
― 6 min read
A new method to improve response speed in language models using selective document processing.
― 8 min read
Exploring bi-reachability challenges in Petri nets enhanced with data values.
― 5 min read
Exploring how AI enhances patent claim drafting efficiency and approval rates.
― 4 min read
KG-FIT combines knowledge graphs with language model insights for richer data representation.
― 7 min read
A study on how language models express and measure their confidence.
― 7 min read
A new algorithm improves code refinement using LLMs more efficiently.
― 6 min read
LLM4EA enhances the efficiency of connecting entities in diverse knowledge graphs.
― 7 min read
A new method enhances reasoning in language models by automating step labeling.
― 6 min read
A new method tackles ethical concerns in language models.
― 5 min read
Zamba is a hybrid language model combining state-space and transformer architectures.
― 6 min read
Exploring the blend of privacy-focused learning and data generation techniques.
― 6 min read
TPO offers a new method to align language models with human preferences efficiently.
― 6 min read
Examining obstacles faced by contributors of low-resourced languages in Ethiopia.
― 5 min read
UltraGist compresses long texts while keeping essential information intact.
― 8 min read
A new framework uses simulated comments to improve fake news detection.
― 6 min read
A method for generating quality training data for language model fine-tuning.
― 7 min read
New techniques enable training large neural networks on consumer-grade hardware with reduced memory.
― 8 min read
DarijaBanking dataset enhances banking systems' understanding of Moroccan Arabic.
― 5 min read
New benchmark aims to improve AI understanding of text and images.
― 7 min read
M-RAG enhances text generation through efficient information retrieval.
― 6 min read
A new method enhances local citation recommendations for researchers.
― 6 min read
Research reveals how large language models respond to various input types.
― 6 min read
Training open-source LLMs enhances optimization modeling for industry applications.
― 7 min read
A new dataset enhances mobile app recommendation systems through conversational exchanges.
― 5 min read
This study examines how storytelling techniques influence emotional connections.
― 6 min read
A new framework improves machine learning from diverse information sources.
― 7 min read
ThReaD improves LLMs' performance on complex tasks through dynamic thread management.
― 5 min read
Harnessing early-exit models for efficient federated learning in ASR systems.
― 8 min read
Improving text generation quality by selecting cleaner examples.
― 7 min read
Exploring the significance and challenges of names in social research.
― 5 min read
A simplified model for effective navigation using natural language instructions.
― 10 min read