AI
Lightweight Adaptive ReduNet via Hyperspherical Manifold Learning
Researchers have proposed LA-ReduNet, a lightweight version of the neural network ReduNet. Unlike tr...
AI
Temporal Validity on Real Software Histories: Eliminating Stale-Fact Errors in Code-Assistant Memory over GitHub Fixes
Researchers have developed a method to eliminate 'stale-fact errors' in code-assistant memory, where...
AI
Identity-Aware Human-Object Interaction Motion Captioning
Researchers have proposed a new task for human-object interaction (HOI) motion captioning that requi...
AI
Vis-Poison: Poisoning Visual Knowledge in Multimodal Retrieval-Augmented Generation
Researchers have developed a new attack method called Vis-Poison that can compromise multimodal larg...
AI
PSK at WMT 2026 MIST: Task-Specialized QLoRA Adapters for Multilingual Summarization and Question Answering
Researchers from Srikar Kashyap Pulipapa's team submitted a system to the WMT 20...
AI
Fuzzy-MoE: Interpretable Regime-Conditioned Expert Routing for Non-Stationary Multivariate Time Series Forecasting
A new AI model called Fuzzy-MoE has been proposed for forecasting non-stationary...
AI
CARD: Diagnosing Belief to Action Routing Failures in Vision Language Models
Researchers have developed a tool called CARD that helps diagnose problems in ho...
AI
Do SpeechLMs Hear Their Own Opinions? Diagnosing and Mitigating Previous-Belief Contamination in Streaming Emotion Understanding
Researchers have found that speech models (SpeechLMs) can be contaminated by the...