作者位次 #2/6ACL 2026Poster· Natural Language Processing
Truth as a Trajectory: What Internal Representations Reveal About Large Language Model Reasoning
Hamed Damirchi · Imezadelajara · Ehsan Abbasnejad · Afshar Shamsi · Zhen Zhang · Javen Qinfeng Shi
摘要
Existing explainability methods for Large Language Models (LLMs) typically treat hidden states as static points in activation space, assuming that correct and incorrect inferences can be separated using representations from an individual layer. However, these activations are saturated with polyse...