AI
Evidence-Consistent Generative Detection under Scenario-Level Distribution Shift
Researchers propose a new framework called ECoG for detecting out-of-distribution (OOD) scenarios in...
AI
Jokes Aside: Measuring the Semantic Distance of Double Meanings
Researchers have developed new metrics to measure the semantic distance between words in jokes, aimi...
AI
When the Feature Pool Goes Algorithmic: Extending Mufwene's Ecology of Language Evolution to LLM-Mediated Exposure
Researchers have proposed a new framework for understanding how language evolves in the presence of ...
AI
RARE: Decoupling Representation Steering from Expert Routing in Mixture-of-Experts Language Models
Researchers have developed a new framework called RARE that allows them to control the behavior of l...
AI
Affective Context Amplifies Sycophancy in LLM Responses
Researchers have found that large language models (LLMs) are more likely to be s...
AI
Benchmarking Patent Drafting from Inventor-Style Disclosures
Researchers have created a new dataset called Dis2Pat to help improve the perfor...
AI
Memory Augmentation Unlocks Efficient Chain-of-Thought Reasoning
Researchers have proposed a new framework called Memory-Augmented Compression to...
AI
Prompt-Model Interaction Reaches the Fixed Points: A deterministic, task-free structural readout -- and the factorizations of it that failed
Researchers have explored how prompts interact with models, finding that the eff...