Internal Flow Signatures for Self-Checking and Refinement in LLMs
This addresses the issue of unfaithful generation in LLMs for users relying on accurate outputs, though it appears incremental as it builds on existing monitoring and refinement techniques.
The paper tackles the problem of large language models generating unfaithful answers by introducing internal flow signatures that audit decision formation from depthwise dynamics, enabling self-checking and targeted refinement without modifying the base model, resulting in actionable localization and low-overhead detection.
Large language models can generate fluent answers that are unfaithful to the provided context, while many safeguards rely on external verification or a separate judge after generation. We introduce \emph{internal flow signatures} that audit decision formation from depthwise dynamics at a fixed inter-block monitoring boundary. The method stabilizes token-wise motion via bias-centered monitoring, then summarizes trajectories in compact \emph{moving} readout-aligned subspaces constructed from the top token and its close competitors within each depth window. Neighboring window frames are aligned by an orthogonal transport, yielding depth-comparable transported step lengths, turning angles, and subspace drift summaries that are invariant to within-window basis choices. A lightweight GRU validator trained on these signatures performs self-checking without modifying the base model. Beyond detection, the validator localizes a culprit depth event and enables a targeted refinement: the model rolls back to the culprit token and clamps an abnormal transported step at the identified block while preserving the orthogonal residual. The resulting pipeline provides actionable localization and low-overhead self-checking from internal decision dynamics. \emph{Code is available at} \texttt{github.com/EavnJeong/Internal-Flow-Signatures-for-Self-Checking-and-Refinement-in-LLMs}.