At the Edge of Understanding: Sparse Autoencoders Trace The Limits of Transformer Generalization

ICLR 2026 Conference Submission21952 Authors

Keywords: robustness, generalization, sparse autoencoders, transformers

Abstract: Pre-trained transformers have demonstrated remarkable generalization abilities, at times extending beyond the scope of their training data. Yet, real-world deployments often face unexpected or adversarial data that diverges from training data distributions. Without explicit mechanisms for handling such shifts, model reliability and safety degrade, urging more disciplined study of out-of-distribution (OOD) settings for transformers. By systematic experiments, we present a mechanistic framework for delineating the precise contours of transformer model robustness. We find that OOD inputs, including subtle typos and jailbreak prompts, drive language models to operate on an increased number of fallacious concepts in their internals. We leverage this device to quantify and understand the degree of distributional shift in prompts, enabling a mechanistically grounded fine-tuning strategy to robustify LLMs. Expanding the very notion of OOD from input data to a model’s private computational processes—a new transformer diagnostic at inference time—is a critical step toward making AI systems safe for deployment across science, business, and government.

Primary Area: interpretability and explainable AI

Submission Number: 21952