Uncertainty-based Debiasing and Unlearning for Decontamination

• URL: http://arxiv.org/abs/2606.23313v1
• Date: Mon, 22 Jun 2026 13:26:29 GMT
• Title: Uncertainty-based Debiasing and Unlearning for Decontamination
• Abstract summary: Uncertainty-Based Decontamination (UBD) estimates per-sample memorization without requiring a contaminated model or knowledge of which samples are contaminated.<n>UBD produces per-sample output distributions substantially closer to those of an uncontaminated model than paraphrasing or choice-permutation baselines.
• Score: 12.593872705071114
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Benchmark-based evaluation is the dominant paradigm for assessing large language model (LLM) capabilities, yet data contamination inflates reported performance and undermines fair comparison. Existing decontamination methods are evaluated solely through aggregate accuracy, which can obscure substantial differences in per-sample model behaviour, and many require access to an uncontaminated model. In this paper, we propose a sample-level evaluation framework for decontamination that complements accuracy-based assessment with distributional distance metrics, measuring how closely a decontaminated model recovers the output distribution of an uncontaminated model on each sample. Building on this framework, we introduce Uncertainty-Based Decontamination (UBD), a family of methods that leverage deep ensembles of the contaminated model to estimate per-sample memorization without requiring a uncontaminated model or knowledge of which samples are contaminated. UBD estimates a per-sample correction scalar from ensemble uncertainty, which is used to construct a debiased target distribution that suppresses the inflated probability mass on correct answers induced by contamination. This target is then used either as a post-hoc output correction (debiasing) or as a soft training signal for parameter update (unlearning). Experiments on MMLU-Pro and MATH-MCQA across multiple LLM backbones demonstrate that UBD produces per-sample output distributions substantially closer to those of an uncontaminated model than paraphrasing or choice-permutation baselines, while preserving model performance on uncontaminated data.

Related papers

• Impute-EM: Native Mixed-State Diffusion Models for Heterogeneous Data Imputation [71.74203645336385]
  Impute-EM alternates between imputing missing entries with the current model and refitting a diffusion backbone on completed data.<n>We characterize the update and show that the observed mask-indexed marginals match the targets at the limit.
  arXiv  Detail & Related papers  (2026-09-14T09:39:13Z)
• Wasserstein Filtering: A Sample Selection Method for Robust Distribution Learning [4.341190001906841]
  Wasserstein Filtering (WF) is a novel sample selection framework that discards a fraction of suspicious samples and estimates the target distribution.<n>WF serves as a highly practical, model-agnostic preprocessing tool.<n>It delivers competitive outlier detection performance and provides substantial downstream benefits for generative modeling under heavy contamination.
  arXiv  Detail & Related papers  (2026-08-13T16:12:28Z)
• Distributionally Faithful Imputation via Positive Semi-Definite Kernel Density Estimation [55.47411235954028]
  We recast imputation under missing completely at random as density estimation from masked observations.<n>We obtain a convex empirical risk problem with closed form marginals, solvable by a Newton interior point method.<n>The resulting PSD Impute model yields both single and multiple imputations from the same fitted density.
  arXiv  Detail & Related papers  (2026-07-08T15:51:05Z)
• A Theoretical Framework for Statistical Evaluability of Generative Models [57.9316356505791]
  We introduce a theoretical framework for evaluating generative models and establish evaluability results for commonly used metrics.<n>We study two categories of metrics: test-based metrics, including integral probability metrics (IPMs) and Rényi divergences.<n>We show that IPMs with respect to any bounded test class can be evaluated from finite samples up to multiplicative and additive approximation errors.<n>In contrast, Rényi and KL divergences are not evaluable from finite samples, as their values can be critically determined by rare events.
  arXiv  Detail & Related papers  (2026-04-07T01:53:59Z)
• Heterogeneous Multisource Transfer Learning via Model Averaging for Positive-Unlabeled Data [2.030810815519794]
  We propose a novel transfer learning framework that integrates information from heterogeneous data sources without direct data sharing.<n>For each source domain type, a tailored logistic regression model is conducted, and knowledge is transferred to the PU target domain through model averaging.<n>Our method outperforms other comparative methods in terms of predictive accuracy and robustness, especially under limited labeled data and heterogeneous environments.
  arXiv  Detail & Related papers  (2025-11-14T03:15:31Z)
• Reference-Specific Unlearning Metrics Can Hide the Truth: A Reality Check [60.77691669644931]
  We propose Functional Alignment for Distributional Equivalence (FADE), a novel metric that measures distributional similarity between unlearned and reference models.<n>We show that FADE captures functional alignment across the entire output distribution, providing a principled assessment of genuine unlearning.<n>These findings expose fundamental gaps in current evaluation practices and demonstrate that FADE provides a more robust foundation for developing and assessing truly effective unlearning methods.
  arXiv  Detail & Related papers  (2025-10-14T20:50:30Z)
• DDB: Diffusion Driven Balancing to Address Spurious Correlations [24.940576844328408]
  Deep neural networks trained with Empirical Risk Minimization often fail to generalize to out-of-distribution samples.<n>We propose a Diffusion Driven Balancing (DDB) technique to generate training samples with text-to-image diffusion models.<n>Our experiments show that our technique achieves better worst-group accuracy than the existing state-of-the-art methods.
  arXiv  Detail & Related papers  (2025-03-21T15:28:22Z)
• Theory on Score-Mismatched Diffusion Models and Zero-Shot Conditional Samplers [49.97755400231656]
  We present the first performance guarantee with explicit dimensional dependencies for general score-mismatched diffusion samplers.<n>We show that score mismatches result in an distributional bias between the target and sampling distributions, proportional to the accumulated mismatch between the target and training distributions.<n>This result can be directly applied to zero-shot conditional samplers for any conditional model, irrespective of measurement noise.
  arXiv  Detail & Related papers  (2024-10-17T16:42:12Z)
• Informed Correctors for Discrete Diffusion Models [27.295990499157814]
  We propose a predictor-corrector sampling scheme for discrete diffusion models.<n>We show that our informed corrector consistently produces superior samples with fewer errors or improved FID scores.<n>Our results underscore the potential of informed correctors for fast and high-fidelity generation using discrete diffusion.
  arXiv  Detail & Related papers  (2024-07-30T23:29:29Z)
• Rejection via Learning Density Ratios [50.91522897152437]
  Classification with rejection emerges as a learning paradigm which allows models to abstain from making predictions.<n>We propose a different distributional perspective, where we seek to find an idealized data distribution which maximizes a pretrained model's performance.<n>Our framework is tested empirically over clean and noisy datasets.
  arXiv  Detail & Related papers  (2024-05-29T01:32:17Z)
• ConStat: Performance-Based Contamination Detection in Large Language Models [7.305342793164905]
  ConStat is a statistical method that reliably detects and quantifies contamination by comparing performance between a primary and reference benchmark relative to a set of reference models. We demonstrate the effectiveness of ConStat in an extensive evaluation of diverse model architectures, benchmarks, and contamination scenarios.
  arXiv  Detail & Related papers  (2024-05-25T15:36:37Z)
• Quantifying the Uncertainty in Model Parameters Using Gaussian Process-Based Markov Chain Monte Carlo: An Application to Cardiac Electrophysiological Models [7.8316005711996235]
  Estimates of patient-specific model parameters are important for personalized modeling. Standard Markov Chain Monte Carlo sampling requires repeated model simulations that are computationally infeasible. A common solution is to replace the simulation model with a computationally-efficient surrogate for a faster sampling.
  arXiv  Detail & Related papers  (2020-06-02T23:48:15Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.