Data Kernel Perspective Space Performance Guarantees for Synthetic Data from Transformer Models
This addresses the data scarcity issue for language technology engineers by offering a foundation for statistical guarantees, though it appears incremental as it builds on existing transformer-based synthetic data generation methods.
The paper tackles the problem of unpredictable synthetic data quality from transformer models by proposing Data Kernel Perspective Space (DKPS) to provide mathematical performance guarantees, showing how it elucidates downstream task performance like neural machine translation or LLMs trained with CPO.
Scarcity of labeled training data remains the long pole in the tent for building performant language technology and generative AI models. Transformer models -- particularly LLMs -- are increasingly being used to mitigate the data scarcity problem via synthetic data generation. However, because the models are black boxes, the properties of the synthetic data are difficult to predict. In practice it is common for language technology engineers to 'fiddle' with the LLM temperature setting and hope that what comes out the other end improves the downstream model. Faced with this uncertainty, here we propose Data Kernel Perspective Space (DKPS) to provide the foundation for mathematical analysis yielding concrete statistical guarantees for the quality of the outputs of transformer models. We first show the mathematical derivation of DKPS and how it provides performance guarantees. Next we show how DKPS performance guarantees can elucidate performance of a downstream task, such as neural machine translation models or LLMs trained using Contrastive Preference Optimization (CPO). Limitations of the current work and future research are also discussed.