# When Does Synthetic Patent Data Help? Volume-Fidelity Trade-offs in Low-Resource Multi-Label Classification

arXiv: https://arxiv.org/abs/2605.24296 · HTML: https://mlsanity.com/paper/2605.24296/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 36 references that are in the map.
- First posted on arXiv 22 May 2026.

## Builds on (most cited first, 22 of 22)

- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [QLoRA: Efficient Finetuning of Quantized LLMs](https://mlsanity.com/paper/2305.14314/) (2023)
- [No Language Left Behind: Scaling Human-Centered Machine Translation](https://mlsanity.com/paper/2207.04672/) (2022)
- [Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models](https://mlsanity.com/paper/2506.05176/) (2025)
- [Class-Balanced Loss Based on Effective Number of Samples](https://mlsanity.com/paper/1901.05555/) (2019)
- [A Diversity-Promoting Objective Function for Neural Conversation Models](https://mlsanity.com/paper/1510.03055/) (2015)
- [EDA: Easy Data Augmentation Techniques for Boosting Performance on Text Classification Tasks](https://mlsanity.com/paper/1901.11196/) (2019)
- [Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference](https://mlsanity.com/paper/2412.13663/) (2024)
- [Texygen: A Benchmarking Platform for Text Generation Models](https://mlsanity.com/paper/1802.01886/) (2018)
- [MAUVE: Measuring the Gap Between Neural Text and Human Text using Divergence Frontiers](https://mlsanity.com/paper/2102.01454/) (2021)
- [Asymmetric Loss For Multi-Label Classification](https://mlsanity.com/paper/2009.14119/) (2020)
- [On LLMs-Driven Synthetic Data Generation, Curation, and Evaluation: A Survey](https://mlsanity.com/paper/2406.15126/) (2024)
- [Distribution-Balanced Loss for Multi-Label Classification in Long-Tailed Datasets](https://mlsanity.com/paper/2007.09654/) (2020)
- [SGM: Sequence Generation Model for Multi-label Classification](https://mlsanity.com/paper/1806.04822/) (2018)
- [LLM2LLM: Boosting LLMs with Novel Iterative Data Enhancement](https://mlsanity.com/paper/2403.15042/) (2024)
- [The Parrot Dilemma: Human-Labeled vs. LLM-augmented Data in Classification Tasks](https://mlsanity.com/paper/2304.13861/) (2023)
- [PromptMix: A Class Boundary Augmentation Method for Large Language Model Distillation](https://mlsanity.com/paper/2310.14192/) (2023)
- [PatentSBERTa: A Deep NLP based Hybrid Model for Patent Distance and Classification using Augmented SBERT](https://mlsanity.com/paper/2103.11933/) (2021)
- [PaECTER: Patent-level Representation Learning using Citation-informed Transformers](https://mlsanity.com/paper/2402.19411/) (2024)
- [SynthTextEval: Synthetic Text Data Generation and Evaluation for High-Stakes Domains](https://mlsanity.com/paper/2507.07229/) (2025)
- [Fine-Tuning Causal LLMs for Text Classification: Embedding-Based vs. Instruction-Based Approaches](https://mlsanity.com/paper/2512.12677/) (2025)
- [Evaluating LLM Prompts for Data Augmentation in Multi-label Classification of Ecological Texts](https://mlsanity.com/paper/2411.14896/) (2024)

## Abstract (arXiv)

The issues that must be considered regarding the utilization of synthetic data generated through LLMs for multilabel patent classification include (i) when the use of such data may help and (ii) why. Indeed, the former part appropriately adjusts for the possibility of improving results by an increase in sample size. The current experiment involves six open-source LLMs (from 3.8B to 12B parameters) for four real-data regimes in classification of 64 WIPO labels of assistive technologies. Both full-synthesis generation, conditioned on the label set, and paraphrasing methods are applied, with each used in combination with three classifier categories. It is shown that the claimed improvements in micro F1 for BERT-for-Patents from 0.120 to 0.702 mainly reflect a volume effect; indeed, replication with replacement in 165 examples produces 0.678. Thus, the improvement over the control is +0.024, while compared to the best baseline (focal loss reweighting) is +0.219. The second crucial point to consider here is that of evolving fidelity scores as the data generation regime varies. For low real-data regimes, the volume effect dominates and the correlation coefficient between maximum mean discrepancy (MMD) and classification performance equals r = +0.95. As more real data is used, the correlation becomes inverted and reaches r = -0.73 at the 1:10 regime (Fisher z = +6.47, p < 0.001, 95% CI on Delta r [ +0.96, +1.00 ]). In terms of a fixed budget allocation, combining real data (about 20-30%) with synthetic (70-80%) outperforms both purely synthetic and purely real strategies. Moreover, a corpus that allows for improvement in classification performance up to +0.58 in raw micro F1 may adversely affect a Jaccard-overlap retrieval proxy. Prompt-family variations for other genres may provide some explanation of the phenomenon, but using the standard-patent filter still decreases nDCG@10 by 26%.