If Influence Functions are the Answer, Then What is the Question?
- URL: http://arxiv.org/abs/2209.05364v1
- Date: Mon, 12 Sep 2022 16:17:43 GMT
- Title: If Influence Functions are the Answer, Then What is the Question?
- Authors: Juhan Bae, Nathan Ng, Alston Lo, Marzyeh Ghassemi, Roger Grosse
- Abstract summary: Influence functions efficiently estimate the effect of removing a single training data point on a model's learned parameters.
While influence estimates align well with leave-one-out retraining for linear models, recent works have shown this alignment is often poor in neural networks.
- Score: 7.873458431535409
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: Influence functions efficiently estimate the effect of removing a single
training data point on a model's learned parameters. While influence estimates
align well with leave-one-out retraining for linear models, recent works have
shown this alignment is often poor in neural networks. In this work, we
investigate the specific factors that cause this discrepancy by decomposing it
into five separate terms. We study the contributions of each term on a variety
of architectures and datasets and how they vary with factors such as network
width and training time. While practical influence function estimates may be a
poor match to leave-one-out retraining for nonlinear networks, we show they are
often a good approximation to a different object we term the proximal Bregman
response function (PBRF). Since the PBRF can still be used to answer many of
the questions motivating influence functions, such as identifying influential
or mislabeled examples, our results suggest that current algorithms for
influence function estimation give more informative results than previous error
analyses would suggest.
Related papers
- Do Influence Functions Work on Large Language Models? [10.463762448166714]
Influence functions aim to quantify the impact of individual training data points on a model's predictions.
We evaluate influence functions across multiple tasks and find that they consistently perform poorly in most settings.
arXiv Detail & Related papers (2024-09-30T06:50:18Z) - Causal Feature Selection via Transfer Entropy [59.999594949050596]
Causal discovery aims to identify causal relationships between features with observational data.
We introduce a new causal feature selection approach that relies on the forward and backward feature selection procedures.
We provide theoretical guarantees on the regression and classification errors for both the exact and the finite-sample cases.
arXiv Detail & Related papers (2023-10-17T08:04:45Z) - Stubborn Lexical Bias in Data and Models [50.79738900885665]
We use a new statistical method to examine whether spurious patterns in data appear in models trained on the data.
We apply an optimization approach to *reweight* the training data, reducing thousands of spurious correlations.
Surprisingly, though this method can successfully reduce lexical biases in the training data, we still find strong evidence of corresponding bias in the trained models.
arXiv Detail & Related papers (2023-06-03T20:12:27Z) - GIF: A General Graph Unlearning Strategy via Influence Function [63.52038638220563]
Graph Influence Function (GIF) is a model-agnostic unlearning method that can efficiently and accurately estimate parameter changes in response to a $epsilon$-mass perturbation in deleted data.
We conduct extensive experiments on four representative GNN models and three benchmark datasets to justify GIF's superiority in terms of unlearning efficacy, model utility, and unlearning efficiency.
arXiv Detail & Related papers (2023-04-06T03:02:54Z) - Inf-CP: A Reliable Channel Pruning based on Channel Influence [4.692400531340393]
One of the most effective methods of channel pruning is to trim on the basis of the importance of each neuron.
Previous works have proposed to trim by considering the statistics of a single layer or a plurality of successive layers of neurons.
We propose to use ensemble learning to train a model for different batches of data.
arXiv Detail & Related papers (2021-12-05T09:30:43Z) - 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) - Influence Functions in Deep Learning Are Fragile [52.31375893260445]
influence functions approximate the effect of samples in test-time predictions.
influence estimates are fairly accurate for shallow networks.
Hessian regularization is important to get highquality influence estimates.
arXiv Detail & Related papers (2020-06-25T18:25:59Z) - Explaining Black Box Predictions and Unveiling Data Artifacts through
Influence Functions [55.660255727031725]
Influence functions explain the decisions of a model by identifying influential training examples.
We conduct a comparison between influence functions and common word-saliency methods on representative tasks.
We develop a new measure based on influence functions that can reveal artifacts in training data.
arXiv Detail & Related papers (2020-05-14T00:45:23Z) - RelatIF: Identifying Explanatory Training Examples via Relative
Influence [13.87851325824883]
We use influence functions to identify relevant training examples that one might hope "explain" the predictions of a machine learning model.
We introduce RelatIF, a new class of criteria for choosing relevant training examples by way of an optimization objective that places a constraint on global influence.
In empirical evaluations, we find that the examples returned by RelatIF are more intuitive when compared to those found using influence functions.
arXiv Detail & Related papers (2020-03-25T20:59:54Z)
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.