Deeper Understanding of Black-box Predictions via Generalized Influence Functions

• URL: http://arxiv.org/abs/2312.05586v2
• Date: Mon, 6 May 2024 07:38:21 GMT
• Title: Deeper Understanding of Black-box Predictions via Generalized Influence Functions
• Authors: Hyeonsu Lyu, Jonggyu Jang, Sehyun Ryu, Hyun Jong Yang,
• Abstract summary: Influence functions (IFs) elucidate how data changes model. Increasing size and non- interpretity in large-scale models make IFs inaccurate. We introduce generalized IFs, precisely estimating target parameters' influence while nullifying nuisance gradient changes on fixed parameters.
• Score: 6.649753747542211
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Influence functions (IFs) elucidate how training data changes model behavior. However, the increasing size and non-convexity in large-scale models make IFs inaccurate. We suspect that the fragility comes from the first-order approximation which may cause nuisance changes in parameters irrelevant to the examined data. However, simply computing influence from the chosen parameters can be misleading, as it fails to nullify the hidden effects of unselected parameters on the analyzed data. Thus, our approach introduces generalized IFs, precisely estimating target parameters' influence while nullifying nuisance gradient changes on fixed parameters. We identify target update parameters closely associated with the input data by the output- and gradient-based parameter selection methods. We verify the generalized IFs with various alternatives of IFs on the class removal and label change tasks. The experiments align with the "less is more" philosophy, demonstrating that updating only 5\% of the model produces more accurate results than other influence functions across all tasks. We believe our proposal works as a foundational tool for optimizing models, conducting data analysis, and enhancing AI interpretability beyond the limitation of IFs. Codes are available at https://github.com/hslyu/GIF.

Related papers

• Influence Functions for Scalable Data Attribution in Diffusion Models [52.92223039302037]
  Diffusion models have led to significant advancements in generative modelling. Yet their widespread adoption poses challenges regarding data attribution and interpretability. In this paper, we aim to help address such challenges by developing an textitinfluence functions framework.
  arXiv  Detail & Related papers  (2024-10-17T17:59:02Z)
• DataInf: Efficiently Estimating Data Influence in LoRA-tuned LLMs and Diffusion Models [31.65198592956842]
  We propose DataInf, an efficient influence approximation method that is practical for large-scale generative AI models. Our theoretical analysis shows that DataInf is particularly well-suited for parameter-efficient fine-tuning techniques such as LoRA. 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.
  arXiv  Detail & Related papers  (2023-10-02T04:59:19Z)
• Class based Influence Functions for Error Detection [12.925739281660938]
  Influence functions (IFs) are unstable when applied to deep networks. We show that IFs are unreliable when the two data points belong to two different classes. Our solution leverages class information to improve the stability of IFs.
  arXiv  Detail & Related papers  (2023-05-02T13:01:39Z)
• A Closer Look at Parameter-Efficient Tuning in Diffusion Models [39.52999446584842]
  Large-scale diffusion models like Stable Diffusion are powerful and find various real-world applications. We investigate parameter-efficient tuning in large diffusion models by inserting small learnable modules.
  arXiv  Detail & Related papers  (2023-03-31T16:23:29Z)
• On the Effectiveness of Parameter-Efficient Fine-Tuning [79.6302606855302]
  Currently, many research works propose to only fine-tune a small portion of the parameters while keeping most of the parameters shared across different tasks. We show that all of the methods are actually sparse fine-tuned models and conduct a novel theoretical analysis of them. Despite the effectiveness of sparsity grounded by our theory, it still remains an open problem of how to choose the tunable parameters.
  arXiv  Detail & Related papers  (2022-11-28T17:41:48Z)
• Leveraging Unlabeled Data to Predict Out-of-Distribution Performance [63.740181251997306]
  Real-world machine learning deployments are characterized by mismatches between the source (training) and target (test) distributions. In this work, we investigate methods for predicting the target domain accuracy using only labeled source data and unlabeled target data. We propose Average Thresholded Confidence (ATC), a practical method that learns a threshold on the model's confidence, predicting accuracy as the fraction of unlabeled examples.
  arXiv  Detail & Related papers  (2022-01-11T23:01:12Z)
• Causality-based Counterfactual Explanation for Classification Models [11.108866104714627]
  We propose a prototype-based counterfactual explanation framework (ProCE) ProCE is capable of preserving the causal relationship underlying the features of the counterfactual data. In addition, we design a novel gradient-free optimization based on the multi-objective genetic algorithm that generates the counterfactual explanations.
  arXiv  Detail & Related papers  (2021-05-03T09:25:59Z)
• FastIF: Scalable Influence Functions for Efficient Model Interpretation and Debugging [112.19994766375231]
  Influence functions approximate the 'influences' of training data-points for test predictions. We present FastIF, a set of simple modifications to influence functions that significantly improves their run-time. Our experiments demonstrate the potential of influence functions in model interpretation and correcting model errors.
  arXiv  Detail & Related papers  (2020-12-31T18:02:34Z)
• Interpreting Robust Optimization via Adversarial Influence Functions [24.937845875059928]
  We introduce the Adversarial Influence Function (AIF) as a tool to investigate the solution produced by robust optimization. To illustrate the usage of AIF, we apply it to study model sensitivity -- a quantity defined to capture the change of prediction losses on the natural data. We use AIF to analyze how model complexity and randomized smoothing affect the model sensitivity with respect to specific models.
  arXiv  Detail & Related papers  (2020-10-03T01:19:10Z)
• Estimating Structural Target Functions using Machine Learning and Influence Functions [103.47897241856603]
  We propose a new framework for statistical machine learning of target functions arising as identifiable functionals from statistical models. This framework is problem- and model-agnostic and can be used to estimate a broad variety of target parameters of interest in applied statistics. We put particular focus on so-called coarsening at random/doubly robust problems with partially unobserved information.
  arXiv  Detail & Related papers  (2020-08-14T16:48:29Z)
• Evaluating Prediction-Time Batch Normalization for Robustness under Covariate Shift [81.74795324629712]
  We call prediction-time batch normalization, which significantly improves model accuracy and calibration under covariate shift. We show that prediction-time batch normalization provides complementary benefits to existing state-of-the-art approaches for improving robustness. The method has mixed results when used alongside pre-training, and does not seem to perform as well under more natural types of dataset shift.
  arXiv  Detail & Related papers  (2020-06-19T05:08:43Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.