Evaluating Superhuman Models with Consistency Checks
If machine learning models were to achieve superhuman abilities at various reasoning or decision-making tasks, how would we go about evaluating such models, given that humans would necessarily be poor proxies for ground truth? In this paper, we propose a framework for evaluating superhuman models via consistency checks. Our premise is that while the correctness of superhuman decisions may be impossible to evaluate, we can still surface mistakes if the model's decisions fail to satisfy certain logical, human-interpretable rules. We instantiate our framework on three tasks where correctness of decisions is hard to evaluate due to either superhuman model abilities, or to otherwise missing ground truth: evaluating chess positions, forecasting future events, and making legal judgments. We show that regardless of a model's (possibly superhuman) performance on these tasks, we can discover logical inconsistencies in decision making. For example: a chess engine assigning opposing valuations to semantically identical boards; GPT-4 forecasting that sports records will evolve non-monotonically over time; or an AI judge assigning bail to a defendant only after we add a felony to their criminal record.
Code (2)
Tasks
Decision MakingMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
What's the Meaning of Superhuman Performance in Today's NLU?
In the last five years, there has been a significant focus in Natural Language Processing (NLP) on developing larger Pretrained Language Models (PLMs) and introducing benchmarks such as SuperGLUE and SQuAD to measure the…
PositionReading ComprehensionHumanly Certifying Superhuman Classifiers
Estimating the performance of a machine learning system is a longstanding challenge in artificial intelligence research. Today, this challenge is especially relevant given the emergence of systems which appear to increas…
Deepchecks: Evaluating Retrieval-Augmented Generation (RAG)
Large Language Models (LLMs) augmented with Retrieval-Augmented Generation (RAG) techniques are revolutionizing applications across multiple domains, such as healthcare, finance, and customer service. Despite their poten…
ADC: Adversarial attacks against object Detection that evade Context consistency checks
Deep Neural Networks (DNNs) have been shown to be vulnerable to adversarial examples, which are slightly perturbed input images which lead DNNs to make wrong predictions. To protect from such examples, various defense st…
Objectobject-detectionObject DetectionSST-EM: Advanced Metrics for Evaluating Semantic, Spatial and Temporal Aspects in Video Editing
Video editing models have advanced significantly, but evaluating their performance remains challenging. Traditional metrics, such as CLIP text and image scores, often fall short: text scores are limited by inadequate tra…
Objectobject-detectionObject DetectionObject Tracking+1