# Exposure Bias as Epistemic Underidentification in Recursive Forecasting

arXiv: https://arxiv.org/abs/2606.12990 · HTML: https://mlsanity.com/paper/2606.12990/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 14 of its 23 references that are in the map.
- First posted on arXiv 11 Jun 2026.

## Builds on (most cited first, 14 of 14)

- [What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?](https://mlsanity.com/paper/1703.04977/) (2017)
- [Aleatoric and Epistemic Uncertainty in Machine Learning: An Introduction to Concepts and Methods](https://mlsanity.com/paper/1910.09457/) (2019)
- [Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks](https://mlsanity.com/paper/1506.03099/) (2015)
- [Sequence Level Training with Recurrent Neural Networks](https://mlsanity.com/paper/1511.06732/) (2015)
- [Professor Forcing: A New Algorithm for Training Recurrent Networks](https://mlsanity.com/paper/1610.09038/) (2016)
- [How (not) to Train your Generative Model: Scheduled Sampling, Likelihood, Adversary?](https://mlsanity.com/paper/1511.05101/) (2015)
- [Approximate information state for approximate planning and reinforcement learning in partially observed systems](https://mlsanity.com/paper/2010.08843/) (2020)
- [Robust Probabilistic Time Series Forecasting](https://mlsanity.com/paper/2202.11910/) (2022)
- [Flipped Classroom: Effective Teaching for Time Series Forecasting](https://mlsanity.com/paper/2210.08959/) (2022)
- [Learning from Predictions: Fusing Training and Autoregressive Inference for Long-Term Spatiotemporal Forecasts](https://mlsanity.com/paper/2302.11101/) (2023)
- [Time-Series Classification for Dynamic Strategies in Multi-Step Forecasting](https://mlsanity.com/paper/2402.08373/) (2024)
- [Stratify: Unifying Multi-Step Forecasting Strategies](https://mlsanity.com/paper/2412.20510/) (2024)
- [Epistemic Error Decomposition for Multi-step Time Series Forecasting: Rethinking Bias-Variance in Recursive and Direct Strategies](https://mlsanity.com/paper/2511.11461/) (2025)
- [Expectations vs. Realities: The Cost of MSE-Optimal Forecasting Under Conditional Uncertainty](https://mlsanity.com/paper/2606.04342/) (2026)

## Built on by (most cited first, 1 of 1)

- [Deployment-Side Adaptiveness in Multi-Horizon Volatility Forecasting](https://mlsanity.com/paper/2606.27688/) (2026)

## Abstract (arXiv)

Recursive multi-step forecasting is usually framed as distribution shift: models are trained on observed histories but deployed on their own predictions. We show this framing is incomplete by proving that, under partial observability or state truncation, recursive rollout is also an epistemic underidentification problem. Even with deterministic latent dynamics, one-step Bayes supervision identifies behavior only on observed contexts and need not identify the deployed recursive predictor once rollout queries self-generated induced states whose correct local targets are not determined by numeric state alone. We formalize this with induced states Z and provenance variables P, and derive a decomposition of induced-state error into teacher-forcing/rollout mismatch, representation--class approximation, and provenance information gaps. Empirically, we show that rollout enters a distinct induced-state regime, that fixed induced states define a distinct local corrective task, and that closed-loop gains arise not only from local adaptation but also from changing the induced states visited during rollout. Using a simple binary provenance encoding, provenance-aware correction can further improve performance, though gains are conditional rather than uniform. These results recast exposure bias as reasoning under self-induced epistemic uncertainty.