Shared Autoregressive Context Can Distort Relationships in Synthetic Data
cs.LG, cs.CL
Submitted: 2026-09-26
Updated: 2026-09-26
Comments: 54 pages, 8 figures, including appendices
Code: https://github.com/bitsandbytes-foundation/bitsandbytes
License: http://creativecommons.org/licenses/by/4.0/
The gist: Large language models can generate several records within one autoregressive completion, making earlier answers available as context for later records.
Terminology
Abstract
Large language models can generate several records within one autoregressive completion, making earlier answers available as context for later records. This paper shows that such shared-completion batching can distort relationships among variables in the resulting synthetic data, using controlled tests on synthetic survey respondents. In a matched experiment on 2,000 European Social Survey profiles, generating ten rather than one respondent per request increases mean absolute error in within-country correlations by 48-58% for Qwen3.8-27B and 114-127% for Llama-3.3-70B-Instruct across three seeds, holding profiles, examples, questions and decoding parameters fixed. The distortion primarily reflects exaggerated relationship strength, while retaining substantial agreement with the human ordering of correlations. Controlled interventions establish answer history as a causal channel: re-pairing the same preceding values, with profiles and marginal distributions fixed, changes correlations among subsequently generated responses. Hiding preceding answers reduces correlation error in the tested settings but worsens marginal accuracy. Exploratory corrections across social-attitude, health and economic data likewise show that lower correlation error can coexist with worse marginal distributions and regression estimates. Request construction is therefore part of the data-generating process, and synthetic-data validity must be evaluated against the analyses the generated data are intended to support.
Sources
- Beyond the Mean: Three-Axis Fidelity for Aligning LLM-Based Survey Simulators from Small Pilot Data
- Efficient Memory Management for Large Language Model Serving with PagedAttention
- Plausible but Not Valid: A Psychometric Audit of LLMs as Synthetic Survey Respondents
- Tastes without distinction: silicon samples and the synthetic construction of tastes
- Stochastic Parrots or Singing in Harmony? Testing Five Leading LLMs for their Ability to Replicate a Human Survey with Synthetic Data
- gpt-oss-120b & gpt-oss-20b Model Card
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks