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, Philip S. Yu
cs.LG
Submitted: 2026-08-05
Comments: 23 pages, 14 tables, 4 figures
License: http://creativecommons.org/licenses/by/4.0/
The gist: Frozen pretrained forecasters often fail in structured, recurring ways that are costly to repair through fine-tuning.
Terminology
Abstract
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 compositional search over the raw input channels, and a large language model (LLM) that proposes named feature combinations, binary flags, and short executable code. A single validation-grounded gate accepts or rejects every candidate regardless of its origin, and a validation-selected corrector applies the accepted features or leaves the forecast unchanged. This source-agnostic pipeline also allows prior feature-engineering systems to be evaluated under identical conditions, making CRAFTER an instrument for attributing forecast improvements to the feature source alone. Across six public datasets and six frozen backbones, CRAFTER surpasses every dedicated feature-engineering system at every feature budget, roughly doubling the improvement achieved by the corrector alone and reducing the error of the weakest backbones by up to 27%. These gains are robust across different LLM backends and persist even when applied on top of fine-tuned backbones.
Sources
- LLM-FE: Automated Feature Engineering for Tabular Data with LLMs as Evolutionary Optimizers
- Chronos-2: From Univariate to Universal Forecasting
- VEST: Automatic Feature Engineering for Forecasting
- Post-Training Corrections for Improved Time-Series Forecasting
- Toto: Time Series Optimized Transformer for Observability
- Learning the Context of Errors: Black-Box Online Adaptation of Time Series Foundation Models
- NSR-Boost: A Neuro-Symbolic Residual Boosting Framework for Industrial Legacy Models
- The Forecast After the Forecast: A Post-Processing Shift in Time Series
- Can Molecular Foundation Models Know What They Don't Know? A Simple Remedy with Preference Optimization
- Moirai 2.0: When Less Is More for Time Series Forecasting
- Large Language Models for Automated Data Science: Introducing CAAFE for Context-Aware Automated Feature Engineering
- AUTOCT: Automating Interpretable Clinical Trial Prediction with LLM Agents
- Parameter-Efficient Transfer Learning for NLP
- LoRA: Low-Rank Adaptation of Large Language Models
- Surrogate Modeling for Explainable Predictive Time Series Corrections
- TS-Memory: Plug-and-Play Memory for Time Series Foundation Models
- ELATE: Evolutionary Language model for Automated Time-series Engineering
- Optimized Feature Generation for Tabular Data via LLMs with Decision Tree Reasoning
- Counterfactual Explanations for Time Series Forecasting
- OpenFE: Automated Feature Generation with Expert-level Performance
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks