Inferential Machine Learning: Towards Human-collaborative Foundation Models 

Neural network driven applications like ChatGPT suffer from hallucinations where they confidently provide inaccurate information. A fundamental reason for this inaccuracy is the feed-forward nature of inductive decisions taken by neural networks. Such decisions are a result of training schemes that do not allow networks to deviate from and creatively abduce reasons. With the advent of foundation models that are adapted across applications and data, humans can directly intervene and prompt foundation models. However, without understanding the operational limits of the underlying networks, human interventions often lead to unfair, inaccurate, hallucinated and unintelligible outputs. These outputs undermine the trust in foundation models, thereby causing roadblocks to their adoption in everyday lives. In this tutorial, we provide a systematic way to analyze and understand human interventions in neural network functionality. Specifically, our insights are the following: 1) decision theory must be abductive rather than deductive or inductive, 2) interventions must be analyzed as a function of the ‘not-taken’ residual interventions, 3) interventions are not always positive and networks must be equipped to detect adversarial interactions. The end goal is to promote a human-AI collaborative environment via inferential machine learning techniques.    

Duration: 2 Hours

Presenters

Goals

The goal of the tutorial is threefold: 1) place human intervenability in neural network applicability as a function of existing applications, 2) provide a mathematical framework to understand the effects of interventions in a neural network’s decision-making process, 3) imbue networks with the ability to differentiate between positive and negative interventions. Detailed subtopics are provided in the Outline section below. 

Relevance to Audience

IEEE Computer Society has historically been at the forefront of advancing research in computational decision making. As neural network-based AI systems transition from academia to everyday life, their vulnerabilities must be understood before acceptance by the public. A key vulnerability is the lack of knowledge regarding a neural network’s operational limits. This vulnerability leads to lack of trust in neural networks by non-expert users. In the past, applications of robustness have served the research community to define trustworthy measures. However, users rarely understand these applications. Recently, prompt-based architectures that allow limited inputs from users during inference stage have gained prominence. Hence, users employ a ‘trial and error’ policy that may exceed the operational limits as designed by researchers during training. Hence, a larger emphasis must be placed regarding machine learning at inference. This is a timely tutorial that emphasizes the trustworthiness of neural network predictions in terms of human-centric measures which aid AI’s applicability. Since IEEE Big Data is attended by stakeholders at all levels, including students, researchers, entrepreneurs, policy makers, and other interested parties in academia and industry, it provides an ideal opportunity to emphasize inferential machine learning for foundation models.  

Target Audience

This tutorial is intended for graduate students, researchers, engineers, and policy makers working in different topics related to trustworthy and regulatory machine learning.  

Audience Prerequisites

Audience are expected to have a basic understanding of neural networks and vision applications including image recognition and detection.  

Content Level

Time spent in the short course regarding content level will roughly be split as follows – 10% beginner, 65% intermediate, and 25% advanced.  

Content Description

Neural networks provide generalizable and task independent representation spaces that have garnered widespread applicability in vision, language, and multimodal understanding applications. The complicated semantics of feature interactions within data has been broken down into a set of non-linear functions, convolution parameters, attention, as well as multi-modal inputs among others. The complexity of these operations has introduced multiple vulnerabilities within neural network architectures. These vulnerabilities include adversarial samples, confidence calibration issues, and catastrophic forgetting among others. Given that AI promises to herald the fourth industrial revolution, it is critical to understand and overcome these vulnerabilities. Doing so requires creating trustworthy neural networks that drive the AI systems. Defining trustworthiness, however, is not trivial. Traditionally, trust is a function of trustworthiness attributes including simple invariance to noise and perturbations. However, these measures are neither applicable nor communicable to casual users at inference. In this tutorial, we provide a human-interventionist approach to understanding neural network interactions that allow AI to function in society. Doing so allows us to state the following: 1) neural networks must know when and what they don’t know, 2) neural networks must be amenable to being intervened upon by humans at any stage of their decision-making process. These statements call for neural networks to be equipped with uncertainty quantification and be intervenable. 

We provide a probabilistic post-hoc analysis of uncertainty, and intervenability. Post-hoc implies that a decision has already been made. In larger context, the goal of uncertainty quantification and interventionist amenability must be to satisfy multiple stakeholders at various levels of expertise. This includes researchers, engineers, policymakers, and everyday users among others. In this tutorial, we expound on a gradient-based methodology that provides a probabilistic framework of analysis without requiring any retraining. Once a neural network is trained, it acts as a knowledge base through which different types of gradients can be used to traverse adversarial, contrastive, and counterfactual representation spaces. We demonstrate the utility these gradients to define uncertainty and intervenability. Several image understanding and robustness applications including anomaly, novelty, adversarial, and out-of-distribution image detection, image quality assessment, and noise recognition experiments among others will be discussed. In this tutorial, we examine the types, visual meanings, and interpretations of robustness as a human-centric measure of the utility of large-scale neural networks.  

Tutorial Outline

Previous tutorials

Some of the previous tutorials along with slides and video links can be accessed at https://alregib.ece.gatech.edu/courses-and-tutorials/

Selected Publications 

[1] M. Prabhushankar, and G. AlRegib. “Counterfactual Gradients-based Quantification of Prediction Trust in Neural Networks.” IEEE 7th International Conference on Multimedia Information Processing and Retrieval (MIPR), San Jose, CA, Aug. 7-9, 2024 (Invited Paper) [Github] [Arxiv]

[2] AlRegib, Ghassan, and Mohit Prabhushankar. “Explanatory Paradigms in Neural Networks: Towards relevant and contextual explanations.” IEEE Signal Processing Magazine 39.4 (2022): 59-72.  

[3] M. Prabhushankar, and G. AlRegib, “Introspective Learning : A Two-Stage Approach for Inference in Neural Networks,” in Advances in Neural Information Processing Systems (NeurIPS), New Orleans, LA,, Nov. 29 – Dec. 1 2022.  

[4] Kwon, G., Prabhushankar, M., Temel, D., & AlRegib, G. (2020). Backpropagated gradient representations for anomaly detection. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXI 16,  Springer International Publishing.  

[5] M. Prabhushankar and G. AlRegib, “Extracting Causal Visual Features for Limited Label Classification,” IEEE International Conference on Image Processing (ICIP), Anchorage, AK, Sept 2021.  

[6] Selvaraju, Ramprasaath R., et al. “Grad-cam: Visual explanations from deep networks via gradient-based localization.” Proceedings of the IEEE international conference on computer vision. 2017.  

[7] Prabhushankar, M., Kwon, G., Temel, D., & AlRegib, G. (2020, October). Contrastive explanations in neural networks. In 2020 IEEE International Conference on Image Processing (ICIP) (pp. 3289-3293). IEEE.  

[8] G. Kwon, M. Prabhushankar, D. Temel, and G. AlRegib, “Novelty Detection Through Model-Based Characterization of Neural Networks,” in IEEE International Conference on Image Processing (ICIP), Abu Dhabi, United Arab Emirates, Oct. 2020.  

[9] J. Lee and G. AlRegib, “Gradients as a Measure of Uncertainty in Neural Networks,” in IEEE International Conference on Image Processing (ICIP), Abu Dhabi, United Arab Emirates, Oct. 2020.  

[10] M. Prabhushankar*, G. Kwon*, D. Temel and G. AIRegib, “Distorted Representation Space Characterization Through Backpropagated Gradients,” 2019 IEEE International Conference on Image Processing (ICIP), Taipei, Taiwan, 2019, pp. 2651-2655. (* : equal contribution, Best Paper Award (top 0.1%))  

[11] J. Lee, M. Prabhushankar, and G. AlRegib, “Gradient-Based Adversarial and Out-of-Distribution Detection,” in International Conference on Machine Learning (ICML) Workshop on New Frontiers in Adversarial Machine Learning, Baltimore, MD, Jul., 2022.