Adaptive Activation Cancellation for Hallucination Mitigation in Large Language Models

Researchers developed Adaptive Activation Cancellation (AAC), a real-time framework that reduces hallucinations in large language models by identifying and suppressing problematic neural activations during inference. The method requires no fine-tuning or external knowledge and preserves model capabilities while improving factual accuracy across multiple model scales including LLaMA 3-8B.

• →AAC treats hallucination-associated neural activations as structured interference and suppresses them using confidence-weighted forward hooks during generation.
• →The method requires no external knowledge, fine-tuning, or additional inference passes, making it practically deployable.
• →Testing across OPT-125M, Phi-3-mini, and LLaMA 3-8B showed consistent accuracy improvements on TruthfulQA and HaluEval benchmarks.
• →The framework preserves model capabilities with exactly 0.0% degradation on WikiText-103 perplexity and MMLU reasoning tasks.
• →AAC achieves 5.94x to 3.5x higher selectivity than baseline methods while simultaneously improving factual accuracy and preserving general capabilities.