Pith. sign in

REVIEW 14 cited by

DataInf: Efficiently Estimating Data Influence in LoRA-tuned LLMs and Diffusion Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2310.00902 v3 pith:5QTBSYOJ submitted 2023-10-02 cs.LG stat.ML

DataInf: Efficiently Estimating Data Influence in LoRA-tuned LLMs and Diffusion Models

classification cs.LG stat.ML
keywords datainfinfluencemodelsdatacomputationalexistingfine-tuningmethod
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Quantifying the impact of training data points is crucial for understanding the outputs of machine learning models and for improving the transparency of the AI pipeline. The influence function is a principled and popular data attribution method, but its computational cost often makes it challenging to use. This issue becomes more pronounced in the setting of large language models and text-to-image models. In this work, we propose DataInf, an efficient influence approximation method that is practical for large-scale generative AI models. Leveraging an easy-to-compute closed-form expression, DataInf outperforms existing influence computation algorithms in terms of computational and memory efficiency. Our theoretical analysis shows that DataInf is particularly well-suited for parameter-efficient fine-tuning techniques such as LoRA. Through systematic empirical evaluations, we show that DataInf accurately approximates influence scores and is orders of magnitude faster than existing methods. In applications to RoBERTa-large, Llama-2-13B-chat, and stable-diffusion-v1.5 models, DataInf effectively identifies the most influential fine-tuning examples better than other approximate influence scores. Moreover, it can help to identify which data points are mislabeled.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 14 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. HERMES: A Multi-Granularity Labeling Substrate for Pre-training Data Mixtures

    cs.LG 2026-07 unverdicted novelty 7.0

    HERMES provides a reusable hierarchical labeling substrate for pre-training data that reveals granularity-specific effects in data mixing rules during model training.

  2. DataDignity: Training Data Attribution for Large Language Models

    cs.AI 2026-05 unverdicted novelty 7.0

    ScoringModel raises mean Recall@10 to 52.2 on the FakeWiki provenance benchmark from 35.0 for the best baseline, winning 41 of 45 model-by-condition comparisons and gaining 15.7 points on jailbreak-style queries.

  3. One Modality to Forget Them All: Enhancing Cross-Modal Unlearning in Vision-Language Models

    cs.CV 2026-07 conditional novelty 6.5

    Cross-modal unlearning transfer in vision-language models is asymmetric, architecture-dependent, and shallow under typographic attacks; influence-guided block selection reduces the measured gap.

  4. A Reference-Free Framework for Evaluating Single-Frame ISP Pipelines

    eess.IV 2026-07 conditional novelty 6.0

    A proxy-reference network trained on synthetic camera pipelines estimates PSNR, SSIM, and LPIPS without a ground-truth reference, with LoRA fine-tuning adapting it to real pipelines.

  5. Quantifying the Agreement Between Data-Influence and Data-Similarity to Understand LLM Behavior

    cs.LG 2026-06 unverdicted novelty 6.0

    Data-similarity and data-influence produce significantly overlapping rankings of training documents for LLM outputs, with asymmetry allowing a favorable cost-accuracy trade-off.

  6. DRIFT: Refining Instruction Data via On-Policy Data Attribution

    cs.LG 2026-06 unverdicted novelty 6.0

    DRIFT applies on-policy influence functions with signed weighting and debiasing to attribute and refine SFT data, raising performance on 7B instruction and reasoning models over prior curation methods.

  7. Variance Reduction for Expectations with Diffusion Teachers

    cs.LG 2026-05 unverdicted novelty 6.0

    CARV amortizes upstream diffusion teacher costs over noise resamples with timestep importance sampling and stratified-inverse-CDF sampling, delivering 2-3x effective compute gains in text-to-3D experiments and order-o...

  8. PRISM: Preference-Aware Influence Function Based Data Selection Method for Efficient Fine-Tuning

    cs.LG 2026-05 unverdicted novelty 6.0

    PRISM weights target examples by the current model's preference to build a better representation for influence-function scoring of training samples in efficient LLM fine-tuning.

  9. PRISM: Preference-Aware Influence Function Based Data Selection Method for Efficient Fine-Tuning

    cs.LG 2026-05 unverdicted novelty 6.0

    PRISM weights target examples by model preference to build an improved direction for influence-based data selection in LLM fine-tuning.

  10. On the Fragility of Data Attribution When Learning Is Distributed

    cs.LG 2026-05 unverdicted novelty 6.0

    A single adversary in distributed training inflates its attribution value via latent optimization on synthetic batches without degrading accuracy or triggering basic defenses.

  11. Dataset Distillation by Influence Matching

    cs.CV 2026-07 reject novelty 5.0

    Inf-Match distills datasets by matching estimated parameter influence of real and synthetic data, reporting SOTA classification and retrieval, but with an unsupported theoretical core.

  12. Variance Reduction for Expectations with Diffusion Teachers

    cs.LG 2026-05 unverdicted novelty 5.0

    CARV introduces a hierarchical Monte Carlo estimator with amortized reuse, importance sampling, and stratification that yields 2-3x effective compute gains on diffusion-teacher pipelines while cutting gradient varianc...

  13. Curvature-Weighted Capacity Allocation: A Minimum Description Length Framework for Layer-Adaptive Large Language Model Optimization

    cs.IT 2026-03 conditional novelty 4.0

    A convex-programming framework with closed-form water-filling solutions converts layer curvature scores into optimal expert allocation and pruning ratios for LLMs, with a quadratic transfer-regret bound.

  14. Curvature-Weighted Capacity Allocation: A Minimum Description Length Framework for Layer-Adaptive Large Language Model Optimization

    cs.IT 2026-03 reject novelty 3.0

    A curvature-based layer gain is converted into optimal allocation and pruning via water-filling and convex penalties, but the experiments substitute the baseline's influence scores and show only modest, mixed gains.