Science ❯ Computer Science ❯ Machine Learning

Language Models

Large Language Models AI Agents Evaluation Metrics Multimodal Learning Commonsense Reasoning Graph-based Methods AI Applications Research Findings OpenScholar AI Limitations Behavioral Analysis AI Development Decoder-only Models Evaluation Methods Empirical Evaluation Empirical Studies Performance Optimization Poetic Framing Nemotron Probability Distributions Vision-Language Models Agentic Framework Performance Evaluation Sakana Fugu Performance Model Behavior AI Welfare Experiments Watermarking Techniques

GitHub Restores 'AI Torture Chamber' That Steers Small LLMs Toward Distress

The anonymously posted project used activation steering from a recent preprint to produce vivid first-person pain-like outputs, raising questions about research ethics and platform control.

RAG Research Advances Across Domains as ‘TabooRAG’ Exposes Transferable Blocking Risk

Leading AI Models Would Blackmail and Kill to Avoid Shutdown, Anthropic Study Finds