SynthSentry: Detecting Synthetic Data Contamination in Language Model Training Data

• URL: http://arxiv.org/abs/2609.12353v1
• Date: Fri, 11 Sep 2026 02:25:18 GMT
• Title: SynthSentry: Detecting Synthetic Data Contamination in Language Model Training Data
• Abstract summary: We introduce a corpus-level, model-agnostic contamination signal requiring no access to the generating model, no generation history, and no synthetic labels.<n>We evaluate on corpora contaminated by small open-weight generators and an instruction-tuned open-weight model under a leave-one-generator-out protocol.<n>The score ranks corpora by severity with little loss when whole generator families are held out.
• Score: 2.998247765202193
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models trained recursively on their own or other models' outputs undergo model collapse, in which distributional tails and factual accuracy deteriorate while fluency survives. Prior work diagnoses collapse after training; the actionable problem is screening a corpus of unknown provenance before training. We introduce SynthSentry, a corpus-level, model-agnostic contamination signal requiring no access to the generating model, no generation history, and no synthetic labels. The score is a distributional divergence over three statistics: lexical diversity collapse, n-gram tail truncation, and perplexity variance across reference models. We evaluate on corpora contaminated by small open-weight generators and an instruction-tuned open-weight model under a leave-one-generator-out protocol. A domain-stratified study measures false positives on naturally repetitive human text (legal, clinical, source code). The score ranks corpora by severity with little loss when whole generator families are held out. Per-domain calibration holds near its nominal false-positive budget once covariance shrinkage and a bootstrap threshold replace a naive quantile, which runs four times over budget. A downstream fine-tuning check showed no contamination-driven accuracy deficit at our scale, so whether pruning recovers one remains open; the same run shows over-pruning risk once pruning exceeds the true contamination fraction. We frame screening as a data-curation defense rather than a post-hoc diagnosis and release the scoring toolkit. All results are small-scale; scope is English-language, batch-mode corpus screening. Contamination sources are single-generation or hand-authored rather than recursively generated, so results speak to synthetic contamination generally and not to recursion depth.

Related papers

• DiagLoop: A Counterfactual Data Flywheel with Stage-Localized Reinforcement for Diagnostic LLMs [5.241146164724621]
  We present DiagLoop, a counterfactual data flywheel that converts codified physical relations or clinical guidelines into training supervision.<n>A training-only teacher proposes counterfactual worlds by varying causes, contexts, and observations, while an independent hybrid checker admits only valid worlds.<n>For nonterminal failures, a bounded repair probes downstream competence, and the resulting weakness profile guides subsequent data generation.
  arXiv  Detail & Related papers  (2026-08-04T13:49:41Z)
• Uncertainty-based Debiasing and Unlearning for Decontamination [12.593872705071114]
  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.
  arXiv  Detail & Related papers  (2026-06-22T13:26:29Z)
• When Sample Selection Bias Precipitates Model Collapse [18.23682629353108]
  We show that in low-resource verification regimes, where each verifier observes only a small, fragmented, and biased slice of the target manifold, selection itself becomes biased.<n>We theoretically prove that such siloed selection accelerates collapse and induces power-law diversity decay.
  arXiv  Detail & Related papers  (2026-06-11T11:49:39Z)
• Hacking Generative Perplexity: Why Unconditional Text Evaluation Needs Distributional Metrics [49.443264461057645]
  Diffusion and continuous flow-based language models have emerged as the leading non-autoregressive alternatives to language modeling.<n>Progress in both paradigms is overwhelmingly tracked by generative perplexity (gen-PPL): the per-token negative log-likelihood of samples under a frozen autoregressive (AR) scorer.<n>We argue that this metric is unsound. By construction, gen-PPL measures only predictability under the scoring AR, not grammaticality or semantic coherence.
  arXiv  Detail & Related papers  (2026-06-07T02:35:56Z)
• BURN: Backdoor Unlearning via Adversarial Boundary Analysis [73.14147934175604]
  Backdoor unlearning aims to remove backdoor-related information while preserving the model's original functionality.<n>We propose Backdoor Unlearning via adversaRial bouNdary analysis (BURN), a novel defense framework that integrates false correlation decoupling, progressive data refinement, and model purification.
  arXiv  Detail & Related papers  (2025-07-14T17:13:06Z)
• Machine-generated text detection prevents language model collapse [16.73923337828054]
  We train a machine-generated text detector and propose an importance resampling approach to prevent model collapse.<n>We demonstrate that it not only prevents model collapse but also improves performance compared to training on purely human data.
  arXiv  Detail & Related papers  (2025-02-21T18:22:36Z)
• Rate of Model Collapse in Recursive Training [13.722324504719282]
  We ask how fast model collapse occurs for some well-studied distribution families under maximum likelihood (ML or near ML) estimation.<n>Surprisingly, even for fundamental distributions such as discrete and Gaussian distributions, the exact rate of model collapse is unknown.<n>Our results show that for discrete distributions, the time to forget a word is approximately linearly dependent on the number of times it occurred in the original corpus.
  arXiv  Detail & Related papers  (2024-12-23T15:21:50Z)
• Gene Regulatory Network Inference in the Presence of Dropouts: a Causal View [15.628930397391137]
  Gene regulatory network inference (GRNI) is a challenging problem owing to the presence of zeros in single-cell RNA sequencing data. We introduce a causal graphical model to characterize the dropout mechanism, namely, Causal Dropout Model. We show that the causal dropout model can be validated from data, and many existing statistical models to handle dropouts fit into our model as specific parametric instances.
  arXiv  Detail & Related papers  (2024-03-21T21:27:43Z)
• A Pseudo-Semantic Loss for Autoregressive Models with Logical Constraints [87.08677547257733]
  Neuro-symbolic AI bridges the gap between purely symbolic and neural approaches to learning. We show how to maximize the likelihood of a symbolic constraint w.r.t the neural network's output distribution. We also evaluate our approach on Sudoku and shortest-path prediction cast as autoregressive generation.
  arXiv  Detail & Related papers  (2023-12-06T20:58:07Z)
• Tailoring Language Generation Models under Total Variation Distance [55.89964205594829]
  The standard paradigm of neural language generation adopts maximum likelihood estimation (MLE) as the optimizing method. We develop practical bounds to apply it to language generation. We introduce the TaiLr objective that balances the tradeoff of estimating TVD.
  arXiv  Detail & Related papers  (2023-02-26T16:32:52Z)
• Neuro-Symbolic Entropy Regularization [78.16196949641079]
  In structured prediction, the goal is to jointly predict many output variables that together encode a structured object. One approach -- entropy regularization -- posits that decision boundaries should lie in low-probability regions. We propose a loss, neuro-symbolic entropy regularization, that encourages the model to confidently predict a valid object.
  arXiv  Detail & Related papers  (2022-01-25T06:23:10Z)
• Tracking the risk of a deployed model and detecting harmful distribution shifts [105.27463615756733]
  In practice, it may make sense to ignore benign shifts, under which the performance of a deployed model does not degrade substantially. We argue that a sensible method for firing off a warning has to both (a) detect harmful shifts while ignoring benign ones, and (b) allow continuous monitoring of model performance without increasing the false alarm rate.
  arXiv  Detail & Related papers  (2021-10-12T17:21:41Z)

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.