6 stories tagged Hallucination
Researchers have developed a way to make AI models better at admitting when they don't know something. This could make AI assistants more reliable in everyday use. The method works without needing to see the model's internal workings, making it useful for commercial AI services.
Researchers discovered that providing one confirmed fact in a multi-step reasoning chain increases LLMs' likelihood of confidently producing wrong answers. This phenomenon, called anchored confabulation, challenges assumptions about how models handle partial evidence.
Researchers introduce a dynamic data curation pipeline to better evaluate how LVLMs handle conflicts between visual and textual evidence. The study underscores the importance of models admitting when they lack knowledge.
SymptomWise introduces a hybrid framework that separates language understanding from diagnostic reasoning to eliminate hallucinations in AI symptom analysis. By combining expert-curated knowledge with deterministic inference, the system ensures traceable and consistent outputs in safety-critical settings.
Researchers propose a novel framework treating LLM hallucinations as output-boundary misclassification errors, introducing a composite abstention architecture. This system combines instruction-based refusal with a structural gate that blocks unsupported claims based on a calculated support deficit score.
Researchers propose a weak supervision framework to detect hallucinations in large language models. This approach enables hallucination detection from internal activations alone at inference time.