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

arXiv: https://arxiv.org/abs/2608.05207 · HTML: https://mlsanity.com/paper/2608.05207/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Financial trading and forecasting agents, one of 16 directions in LLM agents and tool use.
- Builds on 21 of its 37 references that are in the map.
- First posted on arXiv 5 Aug 2026.

## Placed in

- LLM agents and tool use / [Financial trading and forecasting agents](https://mlsanity.com/agents/financial-trading-and-forecasting-agents/)

## Builds on (most cited first, 21 of 21)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Parameter-Efficient Transfer Learning for NLP](https://mlsanity.com/paper/1902.00751/) (2019)
- [Chronos: Learning the Language of Time Series](https://mlsanity.com/paper/2403.07815/) (2024)
- [Large Language Models for Automated Data Science: Introducing CAAFE for Context-Aware Automated Feature Engineering](https://mlsanity.com/paper/2305.03403/) (2023)
- [Forecasting day-ahead electricity prices: A review of state-of-the-art algorithms, best practices and an open-access benchmark](https://mlsanity.com/paper/2008.08004/) (2020)
- [Feature Engineering for Predictive Modeling using Reinforcement Learning](https://mlsanity.com/paper/1709.07150/) (2017)
- [OpenFE: Automated Feature Generation with Expert-level Performance](https://mlsanity.com/paper/2211.12507/) (2022)
- [Optimized Feature Generation for Tabular Data via LLMs with Decision Tree Reasoning](https://mlsanity.com/paper/2406.08527/) (2024)
- [Toto: Time Series Optimized Transformer for Observability](https://mlsanity.com/paper/2407.07874/) (2024)
- [LLM-FE: Automated Feature Engineering for Tabular Data with LLMs as Evolutionary Optimizers](https://mlsanity.com/paper/2503.14434/) (2025)
- [AutoMixer for Improved Multivariate Time-Series Forecasting on Business and IT Observability Data](https://mlsanity.com/paper/2310.20280/) (2023)
- [VEST: Automatic Feature Engineering for Forecasting](https://mlsanity.com/paper/2010.07137/) (2020)
- [The Forecast After the Forecast: A Post-Processing Shift in Time Series](https://mlsanity.com/paper/2601.20280/) (2026)
- [AUTOCT: Automating Interpretable Clinical Trial Prediction with LLM Agents](https://mlsanity.com/paper/2506.04293/) (2025)
- [Learning the Context of Errors: Black-Box Online Adaptation of Time Series Foundation Models](https://mlsanity.com/paper/2606.14222/) (2026)
- [TS-Memory: Plug-and-Play Memory for Time Series Foundation Models](https://mlsanity.com/paper/2602.11550/) (2026)
- [ELATE: Evolutionary Language model for Automated Time-series Engineering](https://mlsanity.com/paper/2508.14667/) (2025)
- [AutoFITS: Automatic Feature Engineering for Irregular Time Series](https://mlsanity.com/paper/2112.14806/) (2021)
- [Can Molecular Foundation Models Know What They Don't Know? A Simple Remedy with Preference Optimization](https://mlsanity.com/paper/2509.25509/) (2025)
- [Surrogate Modeling for Explainable Predictive Time Series Corrections](https://mlsanity.com/paper/2412.19897/) (2024)
- [NSR-Boost: A Neuro-Symbolic Residual Boosting Framework for Industrial Legacy Models](https://mlsanity.com/paper/2601.10457/) (2026)

## Abstract (arXiv)

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.