Groundlens
🔍
A proofreader for RAG answers, as a page and an MCP tool
AI & ML interests
NLP, Responsible AI, Hallucination detection, AI verification
Recent Activity
View all activity
AI-that-works
updated a
Space about 2 months ago
AI-that-works
published a
Space about 2 months ago
AI-that-works
updated a
dataset 3 months ago
AI-that-works
authored 6
papers 5 months ago
Semantic Grounding Index: Geometric Bounds on Context Engagement in RAG Systems
Paper • 2512.13771 • Published
A Geometric Taxonomy of Hallucinations in LLMs
Paper • 2602.13224 • Published
How Transformers Reject Wrong Answers: Rotational Dynamics of Factual Constraint Processing
Paper • 2603.13259 • Published
Loss Given Default Prediction Under Measurement-Induced Mixture Distributions: An Information-Theoretic Approach
Paper • 2511.11596 • Published
Capability Ceilings in Autoregressive Language Models: Empirical Evidence from Knowledge-Intensive Tasks
Paper • 2510.21866 • Published
A non-ergodic framework for understanding emergent capabilities in Large Language Models
Paper • 2501.01638 • Published
AI-that-works
published a
dataset 6 months ago