---
title: Uncertainty and Explainability in Deep Rough Volatility: A Neural Information-Theoretic Posterior Approach
url: https://www.ml-quant.com/papers/arxiv/2609.31570/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-10-23
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: arXiv:2609.31570
source_url: https://arxiv.org/abs/2609.31570
featured: 2026-10-23
citations: unknown
topic: Derivatives & Volatility
---


# Uncertainty and Explainability in Deep Rough Volatility: A Neural Information-Theoretic Posterior Approach

Simulation-based inference learns posterior distributions for rough Heston parameters and propagates them through neural pricers to deliver calibrated uncertainty intervals for exotic options.

- Source: https://arxiv.org/abs/2609.31570
- Identifier: arXiv:2609.31570
- Released: 2026-10-23
- First featured: Quant Letter No. 136 (2026-10-23): https://www.ml-quant.com/issues/2026-10-23/
- Citations (Semantic Scholar): not tracked
- Published in: not yet
- Topic: Derivatives & Volatility
- Authors: Damiano Brigo, Raphaël Huser, Dan Leonte

## Abstract (arXiv, CC0)

Deep learning has substantially accelerated the calibration of complex stochastic-volatility models, but neural point calibration alone does not capture the uncertainty remaining after an implied-volatility (IV) surface has been observed. We develop a simulation-based inference framework for rough Heston (rHeston) calibration that learns the posterior distribution of the model parameters conditional on an IV surface. Using neural ratio estimation, we obtain calibrated posterior samples that can be propagated through heteroscedastic neural surrogate pricers for path-dependent exotic options. The resulting posterior-predictive distributions combine residual parameter uncertainty with conditional surrogate uncertainty and yield uncertainty-aware price intervals. We further introduce Hellinger-SHAP, an information-theoretic explainability method for posterior inference. Rather than attributing a single parameter point estimate, it applies local-background Kernel SHAP to a posterior-information functional measuring contraction from the prior to the posterior. This identifies maturity--moneyness regions associated with posterior information gain for individual rHeston parameters. In a simulation study, posterior-predictive intervals provide calibrated or conservative coverage across forward-start, barrier, and realized-variance claims, while point plug-in prices can be materially unreliable for selected contract regimes. Together, the UQ and XAI analyses provide a transparent framework for uncertainty-aware neural calibration and downstream exotic pricing under the specified prior-predictive model.

## Related

- [Valuation of variable annuities under the Volterra mortality and rough Heston models](https://www.ml-quant.com/papers/arxiv/2604.00472/): The paper discusses how to value variable annuity contracts that offer early surrender options. It uses advanced models and deep learning to find the best strategies for surrendering the contracts, while also providing protection against losses through minimum benefits.
- [Rough Volatility: Fact or Artefact?](https://www.ml-quant.com/papers/arxiv/2203.13820/): Fact or Artifact: The study proposes a new method to estimate the roughness of financial asset volatility, attributing observed roughness to microstructure noise.
- [Constrained deep learning for pricing and hedging european options in incomplete markets](https://www.ml-quant.com/papers/arxiv/2511.20837/): This article discusses a method using deep learning to price and hedge European options in incomplete markets, optimizing risk distribution while handling tricky payoffs.
- [Deep Learning Option Pricing with Market Implied Volatility Surfaces](https://www.ml-quant.com/papers/arxiv/2509.05911/): The study introduces a deep learning framework for pricing options based on market-implied volatility surfaces, providing an efficient and scalable method that improves with more data.
- [Prediction of linear fractional stable motions using codifference, with application to non-Gaussian rough volatility](https://www.ml-quant.com/papers/arxiv/2507.15437/): A new method for predicting future changes in linear fractional stable motion (LFSM) has been proposed, which performs better than the fractional Brownian motion in predicting high-frequency FX rates and volatility time series.
- [Multivariate Rough Volatility](https://www.ml-quant.com/papers/ssrn/5065415/): The article introduces a multivariate version of the Rough Fractional Stochastic Volatility model for analyzing logvolatilities, providing an estimator and confirming its theory through simulation.