Synthetic Data Pretraining

Follow this topic Ask about this topic

Momentum

17 papers in the last four weeks, up 240% on the four weeks before. 0.2% of all new papers.

Jul 13Week of Sep 28

Latest papers 82

Oct 8, 2026cs.LG

Is Real-World Training Data Necessary for Generalist Graph Anomaly Detection?

Generalist graph anomaly detection (GAD) aims to build a foundation model that detects anomalies on arbitrary unseen graphs without retraining or fine-tuning. Sufficient data are essential for foundation model training, yet generalist GAD still faces a data shortage, as real-world anomalous graphs are scarce and costly to collect and annotate. To fill this gap, we propose AG-FORGE, an Anomalous Graph generation Forge for automatic synthesis of anomalous graphs, exploring the feasibility of synthetic data-driven training for generalist GAD. Empirically, we find that synthetic data can achieve performance comparable to real-world training, but fail to push the performance boundary further due to the limited capacity of existing methods. To further unlock model capacity as training data scale up, we develop TS-GGAD, a Topology-Semantic coordinated Generalist GAD that captures complementary topological and semantic anomaly evidence, together with a curriculum learning strategy tailored to large-scale synthetic training. Extensive experiments on 14 real-world datasets demonstrate that TS-GGAD, trained on data generated by AG-FORGE, significantly outperforms state-of-the-art methods.

Yujing Liu, Yixin Liu, Yue Tan +4

School of Information and Communication Technology, Griffith University, Australia · School of Computer Science and Technology, Tongji University, China

Anomaly Detection Graph Foundation Models Synthetic Data Pretraining Synthetic Anomaly Generation

Oct 8, 2026cs.LG

Timer-M1: A Multivariate Time Series Foundation Model via Learning Primitives

We introduce Timer-M1, a pretrained multivariate time series foundation model that learns with primitives for zero-shot forecasting. Across domains, time series share elementary temporal and relational patterns, termed primitives, yet differ in how these primitives manifest and evolve across different contexts. Despite progress in zero-shot and task-general forecasting, existing foundation models may still struggle to generalize to complex real-world scenarios. To this end, we develop a primitive-based data synthesis and pretraining pipeline. The synthesis pipeline generates series with temporal primitives shared across domains and then assembles real and generated series into multivariate samples using relational primitives. Afterwards, samples are organized into episodes by assigning distinct channel roles as target variates, past-only covariates, and known-future covariates, ensuring that the model is optimized on predictable variates using available exogenous information. Technically, Timer-M1 further adapts gated two-dimensional Transformer blocks that dynamically allocate cross-variate attention across layers. Across three large-scale forecasting benchmarks, Timer-M1 ranks first on both FEV and TIME and second on GIFT-Eval among most recent time series foundation models. These results support effective primitive-based pretraining as a route to robust general forecasting technique across domains and task settings.

Haoran Zhang, Haixuan Liu, Xingjian Su +5

School of Software, BNRist, Tsinghua University · ByteDance, Beijing, China

Multivariate Time Series Forecasting Time-Series Foundation Models Zero-Shot Time Series Forecasting Time Series Transformers Synthetic Data Pretraining

Oct 8, 2026cs.LG

Conditional Transfer from Controlled Pretraining Mixtures to Code

Synthetic tasks are increasingly used both as probes of language-model capability and as pretraining data. Both uses are often justified by loss reduction: falling loss is treated as informative, and faster loss reduction with more sampling as evidence that a task is worth sampling. We separate three signals. A task is diagnostic when its loss tracks global pretraining progress; it is teachable when its loss responds to its own token budget; and a data source transfers when including it improves a downstream target. We study controlled pretraining in which 70% of the corpus is fixed general Python and the remaining 30% is a simplex over three source families: OpenCodeInstruct, a curated suite of 12 code-adjacent synthetic tasks, and 15 literature-derived probe tasks. Across a task-budget sweep we detect teachability for 14 of 27 tasks, with a sharp asymmetry between the two synthetic families (10/12 curated versus 4/15 literature-derived). Teachability and downstream transfer give different rankings. On the mixture-simplex edge between the curated suite and OpenCodeInstruct, HumanEval pass@20 after a fixed fine-tuning stage rises from 15.9 at pure curated data to its highest observed value, 22.6, at a mixture that is 75% OpenCodeInstruct, then falls to 19.5 at pure OpenCodeInstruct. Curated synthetic data therefore has conditional value: it contributes as a limited share of a mixture that a target-aligned source still dominates. Finally, a loss-based adaptive scheduler exposes the mismatch between residual loss reducibility and downstream transfer. Across three 60k-step free-ratio runs, Ado drives the OpenCodeInstruct share below 5% within the first 5k steps and to 1.2--1.4% by the end of training, and underperforms its matched fixed-mixture controls by 2.4--11.0 percentage points. Optimizing near-term task-loss reduction moves the mixture away from the region that transfers.

Ohad Rubin

Synthetic Data Pretraining Data Mixture Optimization Synthetic Data Evaluation Code Generation

Oct 7, 2026cs.LG

MotherTree: Meta-learning on synthetic data improves decision tree training

Conventional decision tree algorithms produce effective, transparent models that can be audited, communicated, and deployed independently of the training data, but require learning every new task from scratch. In contrast, tabular foundation models demonstrate that meta-learning from a synthetic prior distribution enables strong in-context prediction for previously unseen tasks, especially in small-sample regimes. However, this approach does not produce a standalone model that can be inspected in isolation. We introduce MotherTree, a tabular transformer that meta-learns decision tree induction: given a training set for a new task, it outputs a hard, axis-aligned decision tree, equivalent in form to classically trained trees, in a single forward pass. MotherTree is pre-trained on a synthetic prior using stochastic gradient descent without requiring reference trees for supervision. On established benchmarks with controlled sample size, the approach is competitive with size-matched trees from common algorithms: recursive partitioning, gradient-based tree learning, globally optimal trees, and distillation from tabular foundation models. Notably, MotherTree consistently improves over from-scratch gradient-based learning and acts as a strong initializer: task-specific tuning of the generated tree outperforms the corresponding from-scratch learner on all benchmarks and sample sizes. These results show that meta-learning can provide effective inductive biases for learning stand-alone, small decision tree classifiers.

Ziyuan Wang, Fredrik D. Johansson

Chalmers University of Technology University of Gothenburg

Decision Tree Learning Meta-Learning Tabular Prediction Synthetic Data Pretraining

Oct 6, 2026cs.CV

HuC-VideoMAE: Human-Centric Video Masked Autoencoding from synthetic data

Modern action recognition models rely on video transformers pretrained on massive collections of web-crawled videos, such as Kinetics-700. However, the use of such data raises ethical concerns, as subjects' consent is typically not obtained. Recent high-quality synthetic video datasets generated from motion-capture data, such as BEDLAM2.0, offer a promising ethical alternative. In this work, we investigate self-supervised pretraining of video transformers on synthetic human-motion datasets. We first show that directly applying the standard VideoMAE masking strategy leads to substantially worse performance than pretraining on Kinetics. To address this limitation, we propose a human-centric masking scheme that leverages body keypoints and person bounding box regions. Our approach encourages the model to focus on the structure and dynamics of human motion during pretraining. Experiments on NTU RGB+D and Toyota-Smarthome demonstrate that our method significantly outperforms standard VideoMAE pretraining on synthetic data, closing 49% of the gap to Kinetics pretraining on NTU RGB+D cross-view-subject without using a single real frame during pretraining. To promote the use of ethical action recognition models, we will publicly release our pretrained models.

Ricardo Pizarro, Roberto Valle, José M. Buenaposada +2

Universidad de Alcal´a, Alcal´a de Henares, Spain · Universidad Polit´ecnica de Madrid, Madrid, Spain · Universidad Rey Juan Carlos, M´ostoles, Spain

Synthetic Data Pretraining Masked Autoencoders Self-Supervised Pre-Training Video Action Recognition Video Representation Learning

Oct 6, 2026cs.LG

TAFFY: A Task-Adaptive Tabular Foundation Model with In-Context Diversity

Recent progress in tabular foundation models suggests that training on synthetic tasks can substantially improve in-context learning capabilities, with overall performance largely depending on how well models can infer task-specific predictive relationships from the available context during inference. In this paper, we introduce TAFFY, a tabular foundation model with an In-Context Diversity Prior and a Task-Conditioned Looped Transformer that strengthen this ability. Specifically, to construct each synthetic pretraining context, the In-Context Diversity Prior samples from multiple related environments derived via controlled interventions and distribution shifts on a shared causal process. This in-context diversity encourages the model to learn a more comprehensive and task-specific representation. Moreover, the Task-Conditioned Looped Transformer iteratively and selectively applies a shared group of Transformer blocks to refine contextual representations, with a task-conditioned gate modulating the final hidden-state update. This enables task-adaptive iterative refinement. Together, these components encourage the model to identify predictive relationships from contextual contrasts during pretraining and dynamically modulate context integration for each task. Across six classification and five regression benchmark datasets, TAFFY attains the lowest average rank.

Zijian Li, Xiangchen Song, Gongxu Luo +7

Carnegie Mellon University · Mohamed bin Zayed University of Artificial Intelligence · Guangdong University of Technology +1

Synthetic Data Pretraining Tabular Foundation Models Recurrent Transformers Tabular ICL

Oct 5, 2026cs.CV

A General Pipeline for Dense Illuminant Estimation via Physically Based Synthetic Data

Illuminant estimation is a fundamental problem in computational photography, as it enables the correction of color shifts induced by varying lighting conditions. While learning-based methods have demonstrated strong performance, their progress is hindered by the limited availability of large-scale datasets with accurate illuminant ground-truth. In this work, we propose a general and reusable pipeline to derive dense illuminant chromaticity maps from physically based 3D-rendered scenes. By repurposing an existing 3D scene collection, our approach enables the systematic generation of pixel-wise illuminant annotations under controlled lighting conditions, effectively lowering the barrier to data acquisition for learning-based illuminant estimation. Using this pipeline, we generate a large-scale synthetic set of 74,321 images, which we employ for pre-training both single- and multi-illuminant estimation models. Extensive experiments with state-of-the-art architectures show that synthetic pre-training consistently improves performance, with gains of up to 28% for single-illuminant estimation and up to 57% for multi-illuminant estimation, particularly in data-scarce regimes. These findings demonstrate that synthetic data generation pipelines offer an effective and scalable solution for the pre-training of illuminant estimation methods.

Luca Cogo, Gianmarco Corti, Simone Bianco +1

Department of Informatics, Systems and Communication, University of Milano – Bicocca, Milan 20126, Italy

Synthetic Data Pretraining Computational Imaging Physically Based Rendering

Oct 2, 2026cs.AI

Geometry Meets Physics: Data-Efficient Pre-Training for Unstructured Neural PDE Solvers

Neural surrogate models for Partial Differential Equations (PDEs) on unstructured 3D geometries are often limited by poor generalization and the high cost of generating large-scale training datasets. Consequently, pre-training on massive datasets of related PDE dynamics has emerged as a critical alternative to enhance the robustness and scalability of these models. However, this strategy is neither compute- nor data-efficient, as it relies on massive pre-computed data that is very costly to generate. In this work, we introduce a disk-data-free pre-training framework tailored to both steady-state and transient regimes. For steady-state problems, we propose a geometry-driven strategy that leverages intrinsic shape descriptors to learn representations of complex 3D domains. For transient problems, we introduce a physics-driven approach based on online generation of synthetic PDE data, enabling scalable pre-training without reliance on expensive datasets. Across multiple experiments, our approach achieves faster convergence, greater data efficiency, and higher accuracy during fine-tuning, particularly under realistic low-data regimes. This methodology provides a practical pathway toward data-efficient neural emulators for large-scale simulations.

Luis Medrano-Navarro, Giacomo Baldan, Qiang Liu +4

Technical University of Munich · University of Stuttgart

Synthetic Data Pretraining Neural PDE Solvers PDE Solving PDE Surrogate Modeling Data-Efficient Learning

Sep 30, 2026cs.CL

Effective Synthetic Data Curation Requires Group-Level Signals

Synthetic data now is essential to LLM training, used to strengthen advanced capabilities such as autonomous and long-horizon task execution. Yet recent work shows that training on it at scale can degrade model generation, making it important to decide what synthetic data is worth training on. While current data curation practices do so with individual-level signals (i.e., estimates of each data sample's training utility in isolation), across pre-training and post-training settings we show that this is insufficient for synthetic data, and that group-level signals (i.e., estimates of utility that account for interactions among data samples) are necessary for effective data curation. First, we show that individual-level signals are blind to how samples jointly affect training: synthetic datasets with different compositions can be indistinguishable under individual-level influence yet differ sharply under group-level influence, and curating by the latter yields better downstream performance, particularly in generative capability. Second, we find that group-level signals matter more as training pipelines become increasingly synthetic: among widely used data curation methods, only those incorporating them improve over baseline, with gains increasing when weights capturing relations among samples are amplified. Finally, we translate these findings into practice -- for model developers under a compute budget, we offer a cheap diagnostic that prioritizes which groups of synthetic data most need group-level estimation, recovering much of the benefit of full group-level scoring at a fraction of the compute cost.

Cathy Jiao, Chenyan Xiong

Language Technologies Institute, Carnegie Mellon University

Synthetic Data Pretraining Training Data Curation Data Valuation Synthetic Data Selection

Sep 30, 2026cs.CL

How Much Is an AI Token Worth? Scaling Laws for Wild AI-Generated Web Text

Web text makes up the majority of pretraining data and is increasingly AI-generated. After applying FineWeb quality filtering, we find that 27.5% of tokens from June 2026 web data are labeled as AI-generated by Pangram, rising to 31.1% by August. Unlike synthetic data or model-collapse setups, this wild AI text comes from many models, is written for human readers, and arrives unlabeled in pretraining corpora. How does AI text in the wild affect language model pretraining? To answer this question, we pretrain 800 language models, varying the ratio of added AI tokens to human tokens, and fit scaling laws to held-out losses on both human and AI-generated text. For data-starved models, adding AI tokens to pretraining data initially lowers loss on human text, but the benefit saturates as more are added and quickly reverses into harm. For models trained on high budgets of human text, AI tokens raise loss almost immediately, while the same number of fresh human tokens keeps lowering it. Scaling laws such as Hoffman et al. (2022) fail to predict this behavior. We propose a new scaling law with separate benefit and harm terms that allows the value of an AI token to change sign while also reducing to Chinchilla in the absence of AI text. When fit on smaller models, our scaling law predicts the effect of AI text on held-out human-text loss for models up to 3.6x larger with 41% lower error than the best existing law over all AI ratios. We recommend filtering AI text when the target is human text, repeating human text before expanding the training dataset with AI-generated web text, and reporting validation loss on human and AI text separately AI text remains valuable when the target is AI text. We release WildAI, an 83B-token corpus with AI, topic, and format labels, all 800 models and code at https://github.com/pangramlabs/WildAI.

Jenna Russell, Ben Glickenhaus, Katherine Thai +4

University of Maryland · Pangram Labs

Language Model Pretraining Synthetic Data Pretraining Language Model Scaling Laws

Sep 30, 2026cs.CL

Synthetic Pre-pretraining Survives Scale, but Not as a Grammatical Prior

Pre-pretraining (PPT) on synthetic non-natural language data improves token efficiency during language model pre-training (PT). Prior work attributes this gain to a grammatical prior, i.e., a structural inductive bias learned during PPT that transfers to natural language grammar. However, PPT has only been tested on models of at most 1B parameters and PT budgets below 2B tokens on predominantly web text. It is unknown whether PPT is effective at larger scales and under more realistic PT data mixtures that combine diverse sources (e.g., code and math). We therefore present a comprehensive study on PPT spanning five PPT tasks, four PT data mixtures, four parameter scales (500M to 7B), and PT budgets of up to 100B tokens. Our results demonstrate that the downstream performance and token efficiency gains of PPT persist at scale, e.g., saving at least 21B PT tokens at the 3B scale. However, in contrast to prior work, we find no consistent evidence that these gains stem from a grammatical prior. Downstream performance does not consistently align with grammatical acceptability across model sizes. Instead, we find that downstream gains arise from PPT tasks that improve long-range retrieval. Finally, PPT performance gains are robust to how PT data mixtures are composed and diminish only when web text is absent. Overall, PPT is a low-cost addition to PT, and future PPT task design should target long-range retrieval rather than natural language grammar.

Atsuki Yamaguchi, Tatsuro Inaba, Joel Niklaus +3

University of Sheffield, UK · Mohamed bin Zayed University of Artificial Intelligence, UAE · Hugging Face +3

Language Model Pretraining Synthetic Data Pretraining Linguistic Bias in Language Models

Sep 29, 2026cs.CL

It's All Training: A Fully Synthetic Single-Stage Recipe for LLMs

Current pre-training datasets are derived from web crawls, with all their issues, and were not designed to support mid- and post-training pipelines--for instance, they contain little explicit reasoning. Thus, many frontier labs have begun to develop their own internal datasets, starting from state-of-the-art models, to augment their pre-training data mix, eg, with reasoning traces to address cold-start problems. While demonstratively effective, none of these datasets are public, and the effect of this so-called synthetic data on knowledge and skill acquisition of language models, including small ones, remains poorly understood. We present SYNTH, the first open-source synthetic corpus derived from 58,698 Wikipedia articles that collapses pre-, mid-, and post-training into a single training stage via structured amplification of curated encyclopedic seeds. We evaluate SYNTH by training a suite of models: a 56M tiny model (Monad), 0.3B-0.6B dense models (Baguettotron), and a 13B / 1B-active MoE. At iso-compute, SYNTH outperforms filtered web data, and our models remain competitive with similarly-sized open-weight baselines. Because SYNTH is back-translated from grounded passages, SYNTH-trained models achieve high factual precision despite 10-140x fewer training tokens, with memorization targeted by the seed corpus. These results show that synthetic datasets, including our SYNTH dataset, are capable of producing competitive generalist models from a fraction of the training data, enabling rapid iteration as the frontier advances. These findings open up possibilities for both generalist models with significantly increased data efficiency, as well as domain-specific models where no instruction or conversational data is available. Finally, we publicly release our SYNTH dataset and the suite of Baguettotron models under a permissive license, thus supporting open-source language model development.

Pierre-Carl Langlais, Pieter Delobelle, Yannick Detrois +7

PleIAs · Sorbonne Center for Artificial Intelligence · Sciences Po Médialab +6

Language Model Pretraining Synthetic Data Pretraining Efficient Language Model Training Synthetic Data Generation for Language Models

Sep 24, 2026cs.AI

Self-Play Pretraining with Zero Data

Advances in language modeling have been driven by scaling pretraining on ever more data. Yet, the training data is still largely curated on the model's behalf. A more general approach to pretraining would let the model learn to generate the data most useful for its own improvement. This would provide an effectively unbounded source of training data, limited by compute rather than human knowledge. We introduce Self-Play Pretraining with Zero Data, an initial proof-of-concept towards realizing this vision. Our procedure casts synthetic data generation as a search over the space of all computable structure, taking inspiration from Solomonoff induction. Starting from random initialization, two models learn in tandem: a generator proposes programs interpreted by a universal Turing machine, generating byte sequences, while a learner autoregressively predicts these byte sequences. The learner is trained with standard cross-entropy, while the generator is trained with reinforcement learning to produce sequences at the frontier of the learner's capabilities, yielding an adaptive curriculum. A universal Turing machine gives us a search space over all computable data-generating processes, imposing little domain-specific structure, and self-play searches over this space for useful training data. We test whether zero-shot performance on natural data improves predictably with self-play compute; this is a clean test of transfer since neither generator nor learner is trained on natural data. Across several natural datasets, zero-shot loss exhibits predictable scaling in compute. The models also exhibit in-context learning, and discover recognizable mathematical sequences during training.

Aditya Cowsik, Kfir Dolev, Michael Y. Li +4

Independent Researcher · Tel Aviv University · Stanford University +1

Synthetic Data Pretraining Synthetic Data Generation for Language Models

Sep 22, 2026cs.CL

TransBERT: A Framework for Synthetic Translation in Domain-Specific Language Modeling

The scarcity of non-English language data in specialized domains significantly limits the development of effective Natural Language Processing (NLP) tools. We present TransBERT, a novel framework for pre-training language models using exclusively synthetically translated text, and introduce TransCorpus, a scalable translation toolkit. Focusing on the life sciences domain in French, our approach demonstrates that state-of-the-art performance on various downstream tasks can be achieved solely by leveraging synthetically translated data. We release the TransCorpus toolkit, the TransCorpus-bio-fr corpus (36.4GB of French life sciences text), TransBERT-bio-fr, its associated pre-trained language model and reproducible code for both pre-training and fine-tuning. Our results highlight the viability of synthetic translation in a high-resource translation direction for building high-quality NLP resources in low-resource language/domain pairs.

Julien Knafou, Luc Mottin, Anaïs Mottaz +2

HES-SO, Geneva, Switzerland · SIB, Swiss Institute of Bioinformatics, Geneva, Switzerland

Synthetic Data Pretraining Language Modeling Low-Resource Language Processing

Sep 22, 2026stat.ML

Learning to Fluctuate: Statistical Foundations for Causal Tabular Pretraining

Causal tabular foundation models amortize effect estimation across synthetic mechanisms, but latent-effect supervision rewards posterior shrinkage rather than encoding the repeated-sample response needed in a fixed deployment population. We introduce fluctuation-supervised pretraining (FSP): each synthetic table is labeled by its average treatment effect plus its efficient influence-function fluctuation; deployment remains a frozen forward pass. Along the path

T_{λ,P}=θ(P)+λP_nψ_P

, we prove an endpoint transition: every fixed

λ<1

retains label ambiguity of order

(1-λ)^2/n

, whereas full fluctuation makes the Gaussian label observable and reduces optimal finite-stratum causal label-prediction risk to order

n^{-2}

. A finite-pretraining bound combines label, network, episode-sampling, and optimization errors; its sampling defect controls fixed-mechanism bias, mean squared error, variance, Gaussian approximation, and, with variance-head accuracy, studentized coverage. Complementary lower bounds separate local

n^{-1}

ATE risk from the

\log N/M

excess risk of generic finite-dictionary episode learning. Experiments trace the learned sampling response. Across 24 nonlinear continuous-covariate cells at trained context lengths, continuous-row FSP lowers checkpoint-mean macro RMSE by 7.0% versus S-learner and wins all 12 weak-overlap cells; validation-selected Summary FSP deploys

11.6\times

faster per table in our warm one-thread benchmark. Under effect shift, matched Raw FSP lowers mean-checkpoint RMSE by 54.2% and teacher defect by 99.0% versus latent-effect supervision, and RMSE by 10.2% versus the released CausalPFN-S checkpoint. Known-effect semisynthesis tests coverage; two randomized-study evaluations show that lower RMSE can coexist with residual attenuation.

Zhiheng Zhang

School of Statistics and Data Science Shanghai University of Finance and Economics

Causal Foundation Models Synthetic Data Pretraining Tabular Foundation Models Causal Effect Estimation Influence Functions

Sep 17, 2026cs.AI

QVAC Genesis III: A Large-Scale, High-Quality Open Synthetic STEM Corpus for Efficient Language Model Pre-Training

High-quality pre-training data is a critical bottleneck for educational and STEM-specific language models targeting edge AI and on-device deployment where token budgets are tightly constrained. While major organizations train ever-larger models on private corpora, the open ecosystem lacks STEM-focused synthetic datasets that deliver high per-token learning value efficiently for small models. To address this gap, we introduce QVAC Genesis III, a 191.43B-token, STEM-focused multi-domain synthetic corpus covering 19 domains across several difficulty levels and different educational styles. QVAC Genesis III is built via a dual generation strategy that performs targeted teacher distillation using a weak edge-scale student model as signal: the student's failures are converted into corrective explanations, while its successes are expanded into contrastive option-level reasoning over all answer choices. We further introduce an LLM-as-a-parser evaluation protocol that extracts final answers from free-form outputs and tracks both accuracy and answer validity. To validate the effectiveness of our QVAC Genesis III data, we conduct controlled from-scratch ablations with 1.7B-parameter models, showing that models trained with QVAC Genesis III consistently outperform both models trained with the open-source synthetic corpus Cosmopedia-v2 and the publicly released Cosmo-1B model across ARC, GPQA Diamond, and MMLU STEM benchmarks, achieving up to +28.57% on ARC-E and +21.35% on ARC-C, while reaching a Valid Answer Rate of up to 99.45%.

Davide Vitabile, N. Ranjan, Akshay Nambiar +2

Tether Data, S.A. de C.V. d.b.a. Tether AI Research

Language Model Pretraining Synthetic Data Pretraining Language Modeling Synthetic Data Generation for Language Models

Sep 16, 2026cs.LG

Preventing Model Collapse: A Fisher-Rao Perspective on the Dynamics of Training with Synthetic Data

Large Language Models (LLMs) are now routinely trained using synthetic data, since high-quality human data has been exhausted by the ever increasing needs of larger and larger models. However, recursive training on synthetic data frequently induces model collapse, a degenerative feedback loop where models progressively forget the true underlying data distribution. Training on a mixture of synthetic and fresh human data is a logical countermeasure and can prevent model collapse. However, it is an open question as to what is the exact minimum required ratio of human-to-synthetic data to maintain training stability. In this paper, we establish rigorous theoretical guarantees on the minimum rate of human data required to prevent model collapse. Although previous work established a formal lower bound for this ratio, such bound can be vacuous for very high dimensions, as the analysis relies on the usual Euclidean metric in R^n and is not adapted to the space of categorical probability distributions. Instead, in this paper we explicitly leverage the information-geometric structure of the probability simplex by analyzing the dynamics of the process under the Fisher-Rao metric. We derive quantitative contraction and invariance bounds that are stable and do not become trivial as the dimensions increase. Thus, we show that the effective required data ratio to prevent model collapse is different than previously implied.

Matteo Marchi, João Pedro Silvestre, Bahman Gharesifard +1

Electrical and Computer Engineering Department, University of California at Los Angeles, Los Angeles, CA 90095 USA · Department of Mathematics and Statistics at Queens University, Kingston, ON, Canada

Synthetic Data Pretraining Model Collapse LLM Training

Sep 16, 2026cs.CR

MiST: Mid-Training LLMs for Cybersecurity

Cybersecurity combines high-stakes analysis with complex technical language, making it an impactful and challenging domain for LLMs. We present MiST (Mid-trained Security Transformer), a suite of 8B and 32B models that achieve strong performance on public cybersecurity benchmarks. We use mid-training as an intermediate adaptation stage between general pre-training and cybersecurity training. Rather than performing continual pre-training over large volumes of raw domain text, we curate a compact, expert-vetted seed corpus, and transform it into high-quality domain-specific synthetic training data. The final MiST checkpoints improve mean cybersecurity accuracy by +13.1 and +8.6 absolute percentage points over the corresponding Qwen baselines for 8B and 32B, respectively, corresponding to relative gains of +27.0% and +15.8%. Ablation results further show that these cybersecurity gains arise in the mid-training and supervised fine-tuning stages through a combination of the synthetic data generation flows. Furthermore, we show that MiST provides a stronger initialization for downstream task-specific fine-tuning adaptation and reinforcement learning.

Oded Ovadia, Elad Ben Zaken, Elad Guttman +1

Dream

Synthetic Data Pretraining Domain-Adaptive Pretraining LLMs for Cybersecurity LLM Training

Sep 15, 2026cs.LG

Procedural Pretraining for Molecular Property Prediction

Molecular property prediction is often limited by the small size of labeled downstream datasets, motivating pretraining on large corpora of unlabeled molecules. In this work, we ask whether useful inductive biases can instead be learned from abstract, procedurally generated data before a model sees any molecular data. We introduce a three-stage training pipeline consisting of procedural pretraining, molecular pretraining on SMILES, and downstream fine-tuning, and evaluate several procedural tasks spanning sequence structure, cellular automata, and graph reasoning. We find that procedural pretraining can improve molecular property prediction even after subsequent molecular pretraining: on Lipophilicity, \textsc{Reverse} reduces test error by 4.8%. For context, the magnitude of this improvement is roughly 90% of the performance difference between our 250K-molecule baseline and the publicly released MoLFormer checkpoint pretrained on approximately 100M molecules. Our analysis shows that the benefit is strongest under downstream data scarcity, depends on the structure of the procedural data rather than only surface-level statistics, and does not increase monotonically with additional procedural training. Instead, transfer typically peaks at an intermediate procedural budget and deteriorates as the model approaches convergence on the procedural task. We further find that, for several tasks, much of the transferable information is localized in the attention layers, while feed-forward layers can contribute to over-specialization. These results show that procedural data can provide transferable structure for molecular learning and offer a complementary route to improving performance when labeled molecular data are limited.

Moritz Friedemann, Zachary Shinnick, Philip Torr +1

University of Oxford · Adelaide University · Slater Labs

Synthetic Data Pretraining Self-Supervised Pre-Training Transfer Learning Molecular Property Prediction

Sep 15, 2026cs.CV

Measuring Annotation Efficiency for Handwritten Devanagari Recognition: Sample-Complexity Curves for Four Pretraining Regimes

To train handwritten text recognition systems we need word images and their corresponding transcriptions, and these transcriptions are produced manually. For a script that can be read by only a small number of specialists, this manual transcription is a limitation, because the trained models are supposed to save the time of those same specialists. A relevant question therefore arises: how many transcriptions are needed before a recogniser becomes useful, and how much of that cost can pretraining remove? In this study the answer is measured directly for handwritten Devanagari. We keep the recogniser, optimiser and evaluation protocol the same and change only the number of real transcribed words used for fine-tuning across nine budgets from 10 to 4,000 and four initialisation regimes, with six seeds at every point. The resulting curves are then converted into annotation-equivalent terms. A CER of 0.50 is reached by supervised synthetic pretraining using only 81 transcribed words, whereas random initialisation requires 355, which gives a label multiplier of 4.40 [3.56, 4.99]. There is a zero-shot reference point as well: with no real transcribed words at all, this pretraining is worth about 136 of them. This advantage gets smaller as the target accuracy improves, and at the most demanding target we measure, it cannot be distinguished from no saving at all. A fourth arm in which only the encoder is transferred separates the effect of the pretraining method from that of transfer scope, and masked image modelling is observed to transfer negatively over a bounded range of budgets. We emphasise that the scarcity in this study is constructed by subsampling a large corpus.

Manglesh Kumar Pandey, Sumit Kumar Banshal

Alliance School of Liberal Arts and Sciences, Alliance University Bengaluru, India · Department of Computer Science and Engineering, Alliance University Bengaluru, India

Synthetic Data Pretraining Optical Character Recognition

Sep 14, 2026cs.SD

Rethinking Procedural Audio Pre-training: Source Scaling and Objective Adaptation

Procedural audio has emerged as a viable source for transferable audio representation learning, but its design principles remain unclear.We revisit two questions: how a procedural source should be scaled, and whether training choices developed on natural audio should transfer unchanged to procedural data.Using a controlled source, we separate scale into formula-class coverage C and within-class rendering diversity I.Experiments with FDSL and AudioMAE show that these two forms of scale provide different benefits and depend on the learning formulation and downstream task. A matched AudioMAE study further shows that procedural audio favors low mask ratios (10%--25%), whereas AudioSet-28K favors 50%--75%. Shared-codebook analysis reveals lower patch diversity and stronger temporal predictability in procedural audio. These results motivate source-aware procedural pre-training, where source scaling and learning configuration are considered jointly.Code is available at https://github.com/Cross-Innovation-Lab/Formula-Bank.

Jiajun Peng, Fengrui Liu, Xinyu Liu +1

School of Data Science, University of Science and Technology of China, Hefei 230026, China · School of Computer Science and Technology, East China Normal University, Shanghai 200062, China · School of Psychology, Shanghai Jiao Tong University, Shanghai 200030, China

Synthetic Data Pretraining Audio Representation Learning Masked Autoencoders

Sep 14, 2026cs.LG

Physics-Guided Synthetic High-Frequency Ultrasound Generation for Skin Layer Segmentation

High-frequency ultrasound (HFUS) enables noninvasive visualization of superficial skin structures, but automated skin-layer analysis is limited by the scarcity of densely annotated data. Existing real HFUS datasets commonly provide annotations for superficial targets such as the epidermis and subepidermal low-echogenic band (SLEB), while dense labels for deeper structures such as dermis, subcutaneous tissue, fascia, and muscle are rarely available. We propose a physics-guided synthetic HFUS generation framework for skin layer segmentation. The framework constructs multilayer acoustic skin phantoms, assigns layer dependent acoustic properties, and uses k-Wave simulation to generate paired synthetic HFUS images, dense layer masks, and simulation metadata. To evaluate whether the generated data provide transferable supervision, we use it for downstream segmentation pretraining and fine-tune the models on real Mendeley HFUS data. Synthetic pretraining followed by real fine-tuning achieved real-domain performance comparable to real-only training and improved mean Dice/IoU in three of four evaluated trainable architectures. These results suggest that physics-guided synthetic HFUS images contain transferable anatomical and textural cues for real-domain skin layer segmentation, although further reduction of the synthetic-real appearance gap is needed to enable greater gains. The code and data are available at: https://github.com/Finn-02/synthetic-hfus-skin-layer-segmentation.

Junkyung ju, Kyungho Yoon, Minwoo Shin

Department of Software, Yonsei University (Mirae Campus), Wonju, 26493, Republic of Korea · School of Mathematics and Computing, Yonsei University, Seoul, 03722, Republic of Korea

Synthetic Data Pretraining Ultrasound Image Segmentation Medical Image Segmentation Medical Image Synthesis Physics-Based Simulation

Sep 10, 2026stat.ML

Learning with Synthetic Data via SGD in High-Dimensional Linear Regression

Synthetic data has become a promising way to scale model training beyond limited human-generated data but it may also induce strong model collapse (Dohmatob et al., 2024), where any fixed fraction of synthetic data prevents model performance from improving under data scaling, leaving a non-vanishing excess risk floor. In this paper, we study how synthetic data affects the generalization of one-pass SGD in high-dimensional linear regression with model shift. We establish finite-sample risk bounds for mixed and two-stage training, separating standard bias and variance from source-mismatch effects, namely fluctuation and persistent drift under mixing and filtered initialization bias under two-stage. These bounds reveal a sharp contrast: mixed training induces strong model collapse, while two-stage training avoids the floor by using synthetic data only in the first stage, showing that collapse is not inevitable under a simple data curriculum. Under a random sketch model, we further obtain scaling laws for both protocols, with tight results for mixed training in the optimization-saturated regime. These laws show that larger models may amplify synthetic-induced degradation under mixing, and quantify how high-quality synthetic pretraining may reduce bias in two-stage training. Finally, we establish an exact finite-sample necessary-and-sufficient condition for two-stage training to strictly outperform real-only training under the same real-data budget and identical real-stage updates. Overall, our results highlight that synthetic data is neither inherently harmful nor beneficial; its effect depends critically on both its quality and the training protocol used to incorporate it.

Jichu li, Difan Zou

Renmin University of China · The University of Hong Kong

Synthetic Data Pretraining Model Collapse Linear Regression

Sep 9, 2026stat.ML

Distillation of Synthetic Data for Time Series Foundation Models

Time series foundation models (TSFMs) are increasingly pre-trained on synthetically generated time series trajectories, where the data generating process is known. Current pre-training recipes are based on loss objectives which compare TSFM outputs to realized future values of each trajectory. We instead propose loss objectives which compare TSFM outputs to the conditional forecast distribution of each trajectory, a procedure we call synthetic data distillation (SDD). SDD corresponds to a Rao-Blackwellization of the training objective, in that it leaves the expectation of stochastic gradients unchanged while provably reducing the covariance of the stochastic gradient under the Loewner partial ordering. We empirically validate SDD on a TSFM model family of sizes from

4

M to

2.5

B parameters, and observe faster convergence of validation loss at every model size: on Gaussian Process data, SDD attains or improves upon the Status Quo loss whilst requiring

10\%-40\%

less training iterations.

Niloy Biswas, Noureddine El Karoui

Meta AI

Synthetic Data Pretraining Time-Series Foundation Models Probabilistic Time Series Forecasting

Sep 3, 2026cs.AI

Xiaomi-TabLDM: A Tabular Foundation Model Technical Report

We introduce Xiaomi-TabLDM, a tabular large data foundation model for classification and regression via in-context learning, which delivers superior prediction accuracy without requiring task-specific fine-tuning. Pretrained exclusively on synthetic data generated from structural causal models (SCMs), our model enables more flexible context utilization and more efficient capacity scaling. i) A new performance standard. Strong regression performance across benchmarks: Xiaomi-TabLDM ranks 1st on OpenML-CTR23 and 2nd on regression across TALENT, TabArena, and BCCO, demonstrating consistently strong regression performance across four complementary benchmark suites. Favorable performance--efficiency trade-off: Xiaomi-TabLDM combines strong predictive performance with substantially lower computational cost. For example, on TabArena regression, it achieves the second-highest Elo while using 82% less training time and 68% less prediction time than the top-ranked TabFM. ii) Large-scale synthetic pretraining. Xiaomi-TabLDM expands the coverage and diversity of synthetic tabular data used for pretraining. We also adopt a three-stage training strategy together with dual-stream feature grouping, lightweight Attention Residual, and sparse Mixture-of-Experts, enabling Xiaomi-TabLDM to learn richer feature interactions and expert specialization across diverse tabular tasks. iii) Test-time scaling. Xiaomi-TabLDM further extends tabular prediction through test-time compute scaling, where allocating additional computation at inference time consistently improves predictive performance over the base model.

Xiaomi-TabLDM Team, :, Penghui Wang +10

Xiaomi

Synthetic Data Pretraining Tabular Foundation Models Test-Time Scaling Tabular ICL Tabular Regression

Sep 3, 2026cs.AI

Synthetic Semantic Supervision for Contrastive Code Representation Learning in Small Transformers: An Empirical Study

General-purpose code embeddings power tools for code search, classification, and retrieval. Compact transformer encoders for code typically rely on either human-written docstrings (labor-intensive and inconsistent) or mined structural signals such as execution traces (setting-specific and costly to collect). We empirically study an alternative: contrastive pretraining of small encoders with synthetically generated natural-language descriptions emphasizing code functionality and intent, paired with code in a dual-encoder framework at training and discarded at inference. We benchmark this approach against pretraining-based baselines, generalist LLMs, and embedding-specific models on eight retrieval, classification, and generation tasks across C, C++, and Java. Synthetic semantic supervision yields statistically significant gains over pretraining baselines of the same inference-time size on five of eight tasks, with parity on two more; once fine-tuned, it matches or exceeds zero-shot models two orders of magnitude larger on classification, and it stays on par with execution-aware supervision at matched pretraining data, suggesting a scalable, effective alternative to existing code-representation paradigms.

Kenneth Paulsen, Florian Tambon, Mike Papadakis +1

University of Luxembourg (SnT) · School of Computing, KAIST

Synthetic Data Pretraining Contrastive Learning Representation Learning Information Retrieval

Aug 31, 2026cs.CL

Synthetic Worlds for Temporal Evaluation and Knowledge Updating in LLMs

Large language models (LLMs) rely on static pretraining corpora, causing their knowledge to become outdated over time. Existing approaches for evaluating knowledge edits either suffer from rapid contamination or rely on counterfactual edits that conflict with rigid existing knowledge. In this work, we propose a synthetic, simulation-driven framework for studying knowledge insertion in LLMs. We introduce {\sc ParallelEvents}, a benchmark of fictional yet realistic future worlds that generates coherent event trajectories for controlled evaluation, avoiding contamination while preserving consistency. Building on this dataset, we develop {\sc Synapse}, a training framework that uses model-generated data to update model parameters via mid-training and instruction tuning. This synthetic pipeline enables scalable knowledge integration without costly human-curated data. Empirically, {\sc Synapse} outperforms existing methods by 14.23%, demonstrating that simulation-based synthetic training leads to robust and coherent knowledge insertions.

Jonathan Zheng, Zirui Shao, Alan Ritter +1

Georgia Institute of Technology · Zhejiang University

Synthetic Data Pretraining Knowledge Editing Temporal Reasoning in Language Models Synthetic Benchmark Generation Synthetic Data Generation for Language Models

Aug 31, 2026cs.CL

REER-PT: Reverse-Engineered Reasoning for Perplexity-Guided Pre-training Data Augmentation

As language-model compute continues to scale, high-quality training data is becoming an increasingly important bottleneck. Conventional next-token prediction supervises what follows a context but leaves the intermediate reasoning behind that continuation implicit. We introduce \textbf{REER-PT}, a scalable framework that extends Reverse-Engineered Reasoning (REER) to raw pre-training data. REER-PT identifies continuations that are difficult to predict but can still be inferred from the preceding context, and inserts concise reasoning annotations that reconstruct the missing connection between context and continuation. Candidate annotations are generated and refined offline, with perplexity serving as the optimization signal. Constraints on length and target leakage filter out unhelpful or trivial annotations. This sparse transformation preserves the source text and remains compatible with standard next-token prediction, avoiding online reasoning rollouts during pre-training. We apply REER-PT to transform a source pre-training corpus into an augmented one. Across augmented-data, original-token, and selected-continuation comparisons, perplexity reductions range from 0.42 to 7.29, and only about 0.05% of annotation 13-grams appear verbatim in the source text. We then train two 680M-parameter models with the same architecture and training configuration on the source and augmented corpora, respectively. The augmented-data model gains up to 2.07 percentage points on several knowledge and reasoning benchmarks. Together, the perplexity analysis indicates improved continuation predictability, while the controlled pre-training experiments suggest that this augmentation can improve model performance without changing the standard pre-training objective.

Haoran Que, Jiajun Shi, Ting Huang +7

Peking University · ByteDance Seed · Work done at ByteDance Seed +1

Language Model Pretraining Synthetic Data Pretraining Synthetic Data Generation for Language Models

Aug 13, 2026cs.LG

Synthetic Persona Pretraining: Alignment from Token Zero

As language-model-based AI is increasingly deployed in autonomous settings, aligning its goals and values with those of humans becomes critical. Today, alignment, and the assistant identity itself, are typically introduced only after pretraining, once behavioral priors are already established. This can make values a thin overlay, rather than deeply rooted, and facilitate subsequent misalignment. Pursuing a different paradigm, we introduce Synthetic Persona Pretraining (SPP), which installs the desired assistant persona from token zero in pretraining. First, we annotate pretraining documents with value-aligned first-person reflections derived from a normative value constitution. Second, we pretrain via the standard cross-entropy loss on standard pretraining documents as well as their reflections, which installs the desired persona among a multitude of other personas. Finally, we post-train on user-assistant dialogue data, which binds this desired persona to the assistant identity, a process we call persona binding. By pretraining models up to 3B parameters on 500B tokens, we show that SPP improves constitution following and jailbreak robustness, and reduces the misalignment rate in out-of-distribution moral dilemmas, while preserving capabilities. Early intervention matters: compared with alignment from token zero, introducing SPP only at the end of pretraining yields weaker constitution adherence, does not shift value priorities, and leads to less aligned choices in dilemmas. This advantage depends on persona binding and, importantly, increases with pretraining budget. Overall, our results show that shaping values early is critical for alignment and establish pretraining-time persona interventions as an effective approach to do so.

Julian Minder, Viktor Moskvoretskii, Raghav Singhal +12

EPFL · SJTU · Northeastern University +8

Synthetic Data Pretraining LLM Alignment Moral Reasoning in Language Models LLM Safety Alignment

Aug 6, 2026cs.CV

Patient Pose Assessment Using a CT-Based Framework for Synthetic Data Generation

An adequate diagnostic quality of radiographs is essential for reliable diagnoses and treatment planning. The patient's pose during radiography is one of the most important factors determining the diagnostic quality. Since patient positioning is difficult and not standardized, an automated AI-based approach using depth images to automatically assess the patient's pose before the radiograph has been taken would be helpful. Due to regulatory hurdles, however, it is difficult in practice to acquire the required depth images and corresponding radiographs. In this paper, we present a framework that can generate such training data synthetically from Computed Tomography scans. We further show that by pretraining on our generated synthetic dataset consisting of 3077 image pairs of upper ankle joints, the pose assessment of real upper ankle joints can be improved by up to 11 percentage points.

Manuel Laufer, Dominik Mairhöfer, Malte Sieren +7

Institute for Neuro- and Bioinformatics, University of Lübeck, Germany · University Medical Center Schleswig-Holstein, Lübeck, Germany · IMAGE Information Systems Europe GmbH, Rostock, Germany +1

Synthetic Data Pretraining Medical Imaging Synthetic Data Generation Medical Image Synthesis