# SCARV: Structure-Constrained Aggregation for Stable Sample Ranking in Redundant NLP Datasets

arXiv: https://arxiv.org/abs/2605.00944 · HTML: https://mlsanity.com/paper/2605.00944/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 33 references that are in the map.
- First posted on arXiv 1 May 2026.

## Builds on (most cited first, 25 of 25)

- [GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding](https://mlsanity.com/paper/1804.07461/) (2018)
- [Character-level Convolutional Networks for Text Classification](https://mlsanity.com/paper/1509.01626/) (2015)
- [Understanding Black-box Predictions via Influence Functions](https://mlsanity.com/paper/1703.04730/) (2017)
- [Deduplicating Training Data Makes Language Models Better](https://mlsanity.com/paper/2107.06499/) (2021)
- [An Empirical Study of Example Forgetting during Deep Neural Network Learning](https://mlsanity.com/paper/1812.05159/) (2018)
- [Data Shapley: Equitable Valuation of Data for Machine Learning](https://mlsanity.com/paper/1904.02868/) (2019)
- [Estimating Training Data Influence by Tracing Gradient Descent](https://mlsanity.com/paper/2002.08484/) (2020)
- [Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamics](https://mlsanity.com/paper/2009.10795/) (2020)
- [Deduplicating Training Data Mitigates Privacy Risks in Language Models](https://mlsanity.com/paper/2202.06539/) (2022)
- [TRAK: Attributing Model Behavior at Scale](https://mlsanity.com/paper/2303.14186/) (2023)
- [Towards Efficient Data Valuation Based on the Shapley Value](https://mlsanity.com/paper/1902.10275/) (2019)
- [Representer Point Selection for Explaining Deep Neural Networks](https://mlsanity.com/paper/1811.09720/) (2018)
- [Beta Shapley: a Unified and Noise-reduced Data Valuation Framework for Machine Learning](https://mlsanity.com/paper/2110.14049/) (2021)
- [Efficient Task-Specific Data Valuation for Nearest Neighbor Algorithms](https://mlsanity.com/paper/1908.08619/) (2019)
- [Data Banzhaf: A Robust Data Valuation Framework for Machine Learning](https://mlsanity.com/paper/2205.15466/) (2022)
- [DataInf: Efficiently Estimating Data Influence in LoRA-tuned LLMs and Diffusion Models](https://mlsanity.com/paper/2310.00902/) (2023)
- [Measuring Forgetting of Memorized Training Examples](https://mlsanity.com/paper/2207.00099/) (2022)
- [Data Valuation using Reinforcement Learning](https://mlsanity.com/paper/1909.11671/) (2019)
- [A Distributional Framework for Data Valuation](https://mlsanity.com/paper/2002.12334/) (2020)
- [CS-Shapley: Class-wise Shapley Values for Data Valuation in Classification](https://mlsanity.com/paper/2211.06800/) (2022)
- [Rethinking Data Shapley for Data Selection Tasks: Misleads and Merits](https://mlsanity.com/paper/2405.03875/) (2024)
- [Threshold KNN-Shapley: A Linear-Time and Privacy-Friendly Approach to Data Valuation](https://mlsanity.com/paper/2308.15709/) (2023)
- [Efficient Data Shapley for Weighted Nearest Neighbor Algorithms](https://mlsanity.com/paper/2401.11103/) (2024)
- [A Versatile Influence Function for Data Attribution with Non-Decomposable Loss](https://mlsanity.com/paper/2412.01335/) (2024)
- [Multiply-Robust Causal Change Attribution](https://mlsanity.com/paper/2404.08839/) (2024)

## Abstract (arXiv)

Sample-level rankings are increasingly used in data-centric NLP for analysis, filtering, debugging, and curation, yet existing pipelines typically score training examples pointwise and rank them as if they were independent. This assumption is fragile in the presence of exact duplicates, near-duplicates, paraphrases, and other redundant structure common in NLP corpora, where stochastic training can make highly similar examples receive unstable relative orderings across random seeds. We study stable sample-level ranking under redundancy and propose textscSCARV, a modular aggregation framework that operates on top of an existing scoring proxy. textscSCARV combines robust multi-seed aggregation with a structure-aware aggregation/allocation step over redundancy clusters. Across synthetic redundancy, naturally mined QQP redundancy, multiple proxy families, several NLP tasks, and end-to-end DistilBERT fine-tuning, textscSCARV substantially improves over bare proxy rankings in global and local stability and yields more reproducible ranking-based decisions such as subset selection and suspicious-example retrieval. Our decomposition and compute-aware frontier sharpen the mechanism: robust multi-seed aggregation is the dominant generic stabilizer, while the structure-aware component adds value mainly under low aggregation budgets or when redundancy clusters are informative, naturally occurring, or sufficiently covered. These results position textscSCARV not as a universal data selector or a universally dominant replacement for seed-only aggregation, but as a stability-oriented aggregation layer for proxy-induced rankings in redundant NLP datasets.