Provably Safe Model Updates

• URL: http://arxiv.org/abs/2512.01899v1
• Date: Mon, 01 Dec 2025 17:19:53 GMT
• Title: Provably Safe Model Updates
• Authors: Leo Elmecker-Plakolm, Pierre Fasterling, Philip Sosnin, Calvin Tsay, Matthew Wicker,
• Abstract summary: We introduce a framework for provably safe model updates.<n>We show that relaxing the problem to parameterized abstract domains (orthotopes, zonotopes) yields a tractable primal-dual formulation.<n>This enables efficient certification of updates - independent of the data or algorithm used - by projecting them onto the safe domain.
• Score: 6.7544474785403885
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Safety-critical environments are inherently dynamic. Distribution shifts, emerging vulnerabilities, and evolving requirements demand continuous updates to machine learning models. Yet even benign parameter updates can have unintended consequences, such as catastrophic forgetting in classical models or alignment drift in foundation models. Existing heuristic approaches (e.g., regularization, parameter isolation) can mitigate these effects but cannot certify that updated models continue to satisfy required performance specifications. We address this problem by introducing a framework for provably safe model updates. Our approach first formalizes the problem as computing the largest locally invariant domain (LID): a connected region in parameter space where all points are certified to satisfy a given specification. While exact maximal LID computation is intractable, we show that relaxing the problem to parameterized abstract domains (orthotopes, zonotopes) yields a tractable primal-dual formulation. This enables efficient certification of updates - independent of the data or algorithm used - by projecting them onto the safe domain. Our formulation further allows computation of multiple approximately optimal LIDs, incorporation of regularization-inspired biases, and use of lookahead data buffers. Across continual learning and foundation model fine-tuning benchmarks, our method matches or exceeds heuristic baselines for avoiding forgetting while providing formal safety guarantees.

Related papers

• Not All Preferences Are Created Equal: Stability-Aware and Gradient-Efficient Alignment for Reasoning Models [52.48582333951919]
  We propose a dynamic framework designed to enhance alignment reliability by maximizing the Signal-to-Noise Ratio of policy updates.<n>SAGE (Stability-Aware Gradient Efficiency) integrates a coarse-grained curriculum mechanism that refreshes candidate pools based on model competence.<n> Experiments on multiple mathematical reasoning benchmarks demonstrate that SAGE significantly accelerates convergence and outperforms static baselines.
  arXiv  Detail & Related papers  (2026-02-01T12:56:10Z)
• Adaptive Defense against Harmful Fine-Tuning for Large Language Models via Bayesian Data Scheduler [67.24175911858312]
  Harmful fine-tuning poses critical safety risks to fine-tuning-as-a-service for large language models.<n>Bayesian Data Scheduler (BDS) is an adaptive tuning-stage defense strategy with no need for attack simulation.<n>BDS learns the posterior distribution of each data point's safety attribute, conditioned on the fine-tuning and alignment datasets.
  arXiv  Detail & Related papers  (2025-10-31T04:49:37Z)
• Rethinking Safety in LLM Fine-tuning: An Optimization Perspective [56.31306558218838]
  We show that poor optimization choices, rather than inherent trade-offs, often cause safety problems, measured as harmful responses to adversarial prompts.<n>We propose a simple exponential moving average (EMA) momentum technique in parameter space that preserves safety performance.<n>Our experiments on the Llama families across multiple datasets demonstrate that safety problems can largely be avoided without specialized interventions.
  arXiv  Detail & Related papers  (2025-08-17T23:46:36Z)
• Principled Input-Output-Conditioned Post-Hoc Uncertainty Estimation for Regression Networks [1.4671424999873808]
  Uncertainty is critical in safety-sensitive applications but is often omitted from off-the-shelf neural networks due to adverse effects on predictive performance.<n>We propose a theoretically grounded framework for post-hoc uncertainty estimation in regression tasks by fitting an auxiliary model to both original inputs and frozen model outputs.
  arXiv  Detail & Related papers  (2025-06-01T09:13:27Z)
• LookAhead Tuning: Safer Language Models via Partial Answer Previews [62.529794567687354]
  Fine-tuning enables large language models to adapt to specific domains, but often compromises their previously established safety alignment.<n>We introduce LookAhead Tuning, a lightweight and effective data-driven approach that preserves safety during fine-tuning.
  arXiv  Detail & Related papers  (2025-03-24T18:11:42Z)
• NLSR: Neuron-Level Safety Realignment of Large Language Models Against Harmful Fine-Tuning [37.024666077902225]
  A handful of malicious data uploaded by users can subtly manipulate the finetuning process, resulting in an alignment-broken model.<n>Existing methods to counteract fine-tuning attacks typically require substantial computational resources.<n>We propose textbfNeuron-textbfLevel textbfSafety textbfRealignment.
  arXiv  Detail & Related papers  (2024-12-17T02:59:04Z)
• SMILE: Zero-Shot Sparse Mixture of Low-Rank Experts Construction From Pre-Trained Foundation Models [85.67096251281191]
  We present an innovative approach to model fusion called zero-shot Sparse MIxture of Low-rank Experts (SMILE) construction. SMILE allows for the upscaling of source models into an MoE model without extra data or further training. We conduct extensive experiments across diverse scenarios, such as image classification and text generation tasks, using full fine-tuning and LoRA fine-tuning.
  arXiv  Detail & Related papers  (2024-08-19T17:32:15Z)
• Data-Driven Distributionally Robust Safety Verification Using Barrier Certificates and Conditional Mean Embeddings [0.24578723416255752]
  We develop scalable formal verification algorithms without shifting the problem to unrealistic assumptions. In a pursuit of developing scalable formal verification algorithms without shifting the problem to unrealistic assumptions, we employ the concept of barrier certificates. We show how to solve the resulting program efficiently using sum-of-squares optimization and a Gaussian process envelope.
  arXiv  Detail & Related papers  (2024-03-15T17:32:02Z)
• All models are local: time to replace external validation with recurrent local validation [10.043347396280009]
  External validation is often recommended to ensure the generalizability of ML models. It neither guarantees generalizability nor equates to a model's clinical usefulness. We submit that external validation is insufficient to establish ML models' safety or utility.
  arXiv  Detail & Related papers  (2023-05-05T00:48:23Z)
• Hyperparameter-free Continuous Learning for Domain Classification in Natural Language Understanding [60.226644697970116]
  Domain classification is the fundamental task in natural language understanding (NLU) Most existing continual learning approaches suffer from low accuracy and performance fluctuation. We propose a hyper parameter-free continual learning model for text data that can stably produce high performance under various environments.
  arXiv  Detail & Related papers  (2022-01-05T02:46:16Z)
• Generalizing Variational Autoencoders with Hierarchical Empirical Bayes [6.273154057349038]
  We present Hierarchical Empirical Bayes Autoencoder (HEBAE), a computationally stable framework for probabilistic generative models. Our key contributions are two-fold. First, we make gains by placing a hierarchical prior over the encoding distribution, enabling us to adaptively balance the trade-off between minimizing the reconstruction loss function and avoiding over-regularization.
  arXiv  Detail & Related papers  (2020-07-20T18:18:39Z)

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.