Science ❯ Cognitive Science

Human-Computer Interaction

User Experience Visual Perception Reasoning Perception User Preferences Language Understanding Behavioral Analysis Language Processing AI Ethics Instruction Following Agentic Behavior Question Answering Systems User Trust AI Reasoning AI Behavior Interactive Systems Trust in AI Information Retrieval Decision Making Machine Learning Multimodal Interaction User Modeling AI in Psychology Emotion Recognition Feedback Systems User Intent Understanding User Studies Mental Mapping AI Perception Studies Neural Data Agent-Based Systems User Engagement User Experience Research Visual Understanding User Research Cross-Modal Interaction Collaboration Explainability Instruction Understanding Reasoning Processes Virtual Environments Communication Mechanisms Action Understanding Distraction in Computing Evaluation Techniques Dialogue Management Trustworthiness Conceptual Understanding Virtual Agents Model Behavior Scene Understanding Bias in AI Visual Cognition Skill Acquisition Skill Assessment User Confidence Engagement Explanation Types Visual Context Model Reliability Natural Language Understanding Robotics Readability User Interaction Spatial Awareness Artificial Intelligence Feature Understanding Dialogue Systems Neuroscience Agentic Frameworks Automation AI Limitations Estimators AI Dependency Spatial Reasoning Intuitive Reasoning Social Intelligence Interactive Learning Subjective Perception User Experience Design Model Interpretation User Feedback Visual Processing User Behavior Emotional Intelligence Embodied Cognition AI Systems Natural Language Processing Artificial Intelligence Research AI Evaluation Turing Test Benchmarking Multimodal Systems Task Automation Trustworthy AI AI Performance

Researchers Propose Experimental Fixes for Key Failures in Retrieval‑Augmented Generation

August 25 arXiv papers present targeted methods to recover multi‑step evidence, cut wrong geographic confidence, and slim long‑video indexes while still needing independent replication for production use.

Three Papers Offer Data, Inference and Safety Fixes for Fragile VLA Robot Models

Study Finds Transformers Collapse on Classic Stroop Test

Study Finds GPT‑4.5 Can Be Mistaken for a Human in Short Text Chats

Study Finds Brief AI Assistance Lifts Performance Then Undercuts Persistence

ArXiv Preprint Says Prompts Drive Many LVLM Hallucinations

Karpathy’s File-Based LLM Wiki Spreads as Developers Build Obsidian Second Brains

New RAG Papers Tout Structured, Chunk-Free and Hybrid Retrieval Gains

New ArXiv RAG Papers Tackle Lean Stacks, Adaptive Retrieval, and Knowledge-Graph QA

Two New RAG Preprints Propose Theme-Aligned Hypergraphs and Hierarchical Agents

RAG Research Enters Next Phase With Hybrid Retrieval, Dynamic Graphs, and Context Valuation

New RAG Preprints Propose Retrieval, Reasoning, and Graph Methods to Curb Hallucinations

New Studies Advance Reliability‑Aware, Graph‑Based and Monitored RAG as Real‑World Test Flags Failures