arxiv-2608-05207 SRC-0594 · paper

When Do Corrective Features Help? An Agent for Corrective Feature Discovery on Black-Box Forecasters

Fangxin Wang, Ziyi Zhang, Diyi Zhuang, Langzhou He, Shiyu Wang, Baichuan Mo

Created: 2026-08-05 · Ingested: 2026-10-24

Frozen pretrained forecasters often fail in structured, recurring ways that are costly to repair through fine-tuning. We study corrective feature discovery: mining interpretable features of a frozen forecaster's residual to drive a lightweight post-hoc corrector. Prior automated feature engineering models the data-generating process; corrective features instead model the model-failure process. We present CRAFTER (Corrective Residual Agent with Feature-based Temporal Exploration and Reasoning), which keeps the backbone frozen and mines its residual with two complementary generators: a compositi

Referenced by

Claims in this catalog that draw on this source, and whether as support or counterpoint.

supportsWhen a frozen pretrained time-series forecaster leaves structured residual error, having an LLM propose named semantic features of that residual (e.g. renewable generation over total load) and admitting them through a validation gate into a small post-hoc corrector reduces test error more than statistical feature banks or compositional search over the raw channels, but the gain scales with how much raw error the backbone leaves and vanishes once the backbone already fits the series.