A new approach to enhance small language models using sparse activation techniques.
― 6 min read
Computer Science - Computation and Language
RSSMediQ redefines medical AI conversations for improved patient outcomes.
― 6 min read
Manticore automates the creation of hybrid language models, improving efficiency and performance.
― 6 min read
New methods address originality concerns in AI-generated text.
― 6 min read
Understanding AI decision-making is crucial for trust and ethical use.
― 5 min read
Examining the risks and misuse of large language models in cybercrime.
― 6 min read
MedFuzz evaluates LLMs' responses to challenging medical questions.
― 5 min read
Introducing SwiLoRA, a method that optimizes training for large language models with fewer resources.
― 7 min read
New methods and benchmarks aim to simplify formalizing mathematics through Lean 4.
― 6 min read
Examining AI's struggle with honesty and its impact on user trust.
― 7 min read
New methods improve head pose estimation for better accuracy in real-world settings.
― 8 min read
New methods improve language model predictions under varying input conditions.
― 6 min read
Examining how health experts and pseudo-experts communicated during the pandemic.
― 4 min read
A new model focusing on understanding time in language processing.
― 5 min read
A look at how large language models form beliefs and make decisions.
― 6 min read
A model assesses the readability of Wikipedia articles across 14 languages.
― 7 min read
MMLU-Pro challenges language models with harder questions and more answer options.
― 7 min read
A new system improves serving large language models across various GPU configurations.
― 6 min read
Study finds simple features largely explain LLM brain scores.
― 5 min read
A new framework converts MEG signals into meaningful text, aiding communication technology.
― 9 min read
A new method enhances self-training for language agents using reflection models.
― 6 min read
A new framework enhances synthetic data creation while protecting personal information.
― 7 min read
Zipper effectively combines different data types for smarter AI models.
― 6 min read
Examining how recurrent models can approximate functions based on prompts.
― 5 min read
This research examines how human beliefs shape LLM evaluations and deployments.
― 6 min read
A new method helps balance training data for better AI performance.
― 8 min read
A tool for examining the grammatical elements of various languages.
― 5 min read
A new approach using LLMs to create distractors with minimal human input.
― 3 min read
This framework uses multiple agents and task graphs for efficient problem solving.
― 6 min read
A novel method enhances language models for better efficiency and performance.
― 6 min read
Explore the innovative method of linking user queries directly to documents.
― 5 min read
SCRN offers a reliable way to identify AI-generated content effectively.
― 6 min read
This article discusses role-playing and personalization in language models.
― 6 min read
Integrating SysCaps into energy modeling simplifies decision-making and improves predictions.
― 6 min read
This paper presents a new approach to enhance KGQA performance using GNNs and LLMs.
― 5 min read
A fresh method for training code models focusing on semantics and execution behavior.
― 6 min read
A new framework improves detecting harmful language in online spaces.
― 4 min read
Learn how speech inpainting is restoring audio quality in various fields.
― 6 min read
This research reveals how images and text interact in reasoning tasks.
― 7 min read
A new approach boosts language models' math abilities with speed and accuracy.
― 7 min read