paper-with-me

Papers

Efficiently Controlling Multiple Risks with Pareto Testing

2022-10-14 · Bracha Laufer-Goldshtein, Adam Fisch, Regina Barzilay, Tommi Jaakkola

Machine learning applications frequently come with multiple diverse objectives and constraints that can change over time. Accordingly, trained models can be tuned with sets of hyper-parameters that affect their predictive behavior (e.g., their run-time efficiency versus error rate). As the number of constraints and hyper-parameter dimensions grow, naively selected settings may lead to sub-optimal and/or unreliable results. We develop an efficient method for calibrating models such that their predictions provably satisfy multiple explicit and simultaneous statistical guarantees (e.g., upper-bounded error rates), while also optimizing any number of additional, unconstrained objectives (e.g., total run-time cost). Building on recent results in distribution-free, finite-sample risk control for general losses, we propose Pareto Testing: a two-stage process which combines multi-objective optimization with multiple hypothesis testing. The optimization stage constructs a set of promising combinations on the Pareto frontier. We then apply statistical testing to this frontier only to identify configurations that have (i) high utility with respect to our objectives, and (ii) guaranteed risk levels with respect to our constraints, with specifiable high probability. We demonstrate the effectiveness of our approach to reliably accelerate the execution of large-scale Transformer models in natural language processing (NLP) applications. In particular, we show how Pareto Testing can be used to dynamically configure multiple inter-dependent model attributes -- including the number of layers computed before exiting, number of attention heads pruned, or number of text tokens considered -- to simultaneously control and optimize various accuracy and cost metrics.

📄 PDF Abstract BibTeX arXiv:2210.07913

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

Multi-Head Attention 설명 없음
Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Label Smoothing Label Smoothing is a regularization technique that introduces noise for the labels. This accounts for the fact that datasets may have mistakes in them, so maximizing the…
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
Residual Connection 설명 없음

Similar Papers 제목 키워드 기반

Risk-Controlling Model Selection via Guided Bayesian Optimization

2023-12-04 · Bracha Laufer-Goldshtein, Adam Fisch, Regina Barzilay, Tommi Jaakkola

Adjustable hyperparameters of machine learning models typically impact various key trade-offs such as accuracy, fairness, robustness, or inference cost. Our goal in this paper is to find a configuration that adheres to u…

Bayesian OptimizationFairnessModel Selection

TESTNAV: Pareto-Guided Search for Compositional Robustness Testing

2026-08-20 · Arooj Arif, Tobias Hartung, Elena Botoeva, Alexandros Koliousis arxiv

Deep learning models remain vulnerable to real-world input perturbations, especially when multiple corruptions co-occur in the same input (e.g., brightness shifts and motion blur). Compositional testing reveals these int…

Code Generation

FUPareto: Bridging the Forgetting-Utility Gap in Federated Unlearning via Pareto Augmented Optimization

2026-02-02 · Zeyan Wang, Zhengmao Liu, Yongxin Cai, Chi Li 외 arxiv

Federated Unlearning (FU) aims to efficiently remove the influence of specific client data from a federated model while preserving utility for the remaining clients. However, three key challenges remain: (1) existing unl…

Evolutionary Preference Sampling for Pareto Set Learning

2024-04-12 · Rongguang Ye, Longcan Chen, Jinyuan Zhang, Hisao Ishibuchi

Recently, Pareto Set Learning (PSL) has been proposed for learning the entire Pareto set using a neural network. PSL employs preference vectors to scalarize multiple objectives, facilitating the learning of mappings from…

Evolutionary Algorithms

Max-Rank: Efficient Multiple Testing for Conformal Prediction

2023-11-17 · Alexander Timans, Christoph-Nikolas Straehle, Kaspar Sakmann, Christian A. Naesseth 외

Multiple hypothesis testing (MHT) frequently arises in scientific inquiries, and concurrent testing of multiple hypotheses inflates the risk of Type-I errors or false positives, rendering MHT corrections essential. This …

Conformal PredictionPredictionUncertainty Quantification