Canada

Text-to-Image models such as Stable Diffusion have shown impressive image generation synthesis, thanks to the utilization of large-scale datasets. However, these datasets may contain sexually explicit, copyrighted, or undesirable content, which allows the model to directly generate them. Given that retraining these large models on individual concept deletion requests is infeasible, fine-tuning algorithms have been developed to tackle concept erasing in diffusion models. While these algorithms yield good concept erasure, they all present one of the following issues: 1) the corrupted feature space yields synthesis of disintegrated objects, 2) the initially synthesized content undergoes a divergence in both spatial structure and semantics in the generated images, and 3) sub-optimal training updates heighten the model&#39;s susceptibility to utility harm. These issues severely degrade the original utility of generative models. In this work, we present a new approach that solves all of these challenges. We take inspiration from the concept of classifier guidance and propose a surgical update on the classifier guidance term while constraining the drift of the unconditional score term. Furthermore, our algorithm empowers the user to select an alternative to the erasing concept, allowing for more controllability. Our experimental results show that our algorithm not only erases the target concept effectively but also preserves the model’s generation capability.

AAAI 2024

All but One: Surgical Erasing with Model Preservation in Text-to-Image Diffusion Models

text-to-image generative models

concept ablation

diffusion models

We are pleased to announce the Thirty-Eighth AAAI Conference on Artificial Intelligence (AAAI-24), which will be held in Vancouver, British Columbia at the Vancouver Convention Centre – West Building from 20-27 February, 2024.

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-24 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.

We expect for AAAI-24 to be an in-person conference – one author of all accepted papers will be expected to present work in person unless there are exceptional circumstances that prevent this.<br><br><br><br>

In order to access the AAAI-24 event page you need to register [here](https://aaai.org/aaai-conference/registration/)

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. 

Text-to-Image models such as Stable Diffusion have shown impressive image generation synthesis, thanks to the utilization of large-scale datasets. However, these datasets may contain sexually explicit, copyrighted, or undesirable content, which allows the model to directly generate them. Given that retraining these large models on individual concept deletion requests is infeasible, fine-tuning algorithms have been developed to tackle concept erasing in diffusion models. While these algorithms yield good concept erasure, they all present one of the following issues: 1) the corrupted feature space yields synthesis of disintegrated objects, 2) the initially synthesized content undergoes a divergence in both spatial structure and semantics in the generated images, and 3) sub-optimal training updates heighten the model's susceptibility to utility harm. These issues severely degrade the original utility of generative models. In this work, we present a new approach that solves all of these challenges. We take inspiration from the concept of classifier guidance and propose a surgical update on the classifier guidance term while constraining the drift of the unconditional score term. Furthermore, our algorithm empowers the user to select an alternative to the erasing concept, allowing for more controllability. Our experimental results show that our algorithm not only erases the target concept effectively but also preserves the model’s generation capability.

technical paper

Artificial Intelligence is undoubtedly becoming pervasive in everyday life of everyone.
In this setting, developing correct AI conception since childhood is not only a need to 
be addressed in educational curricula, but is also a children right.

Accordingly, several initiatives at national and international levels aim at promoting AI
and emerging technology literacy, supported also by a proliferation in the literature 
of learning courses covering a variety of topics, learning objectives and targeted ages.
Schools are therefore pushed to introduce innovative activities for children in their
curricula.

In this paper, we report the results of a case study where we tested the contribution 
of an AI block-based course in developing computational thinking, and human 
and AI minds understanding in fifth and sixth grade children.

Does Any AI-Based Activity Contribute to Develop AI Conception? A Case Study with Italian Fifth and Sixth Grade Classes

Question generation (QG) is a natural language processing task with an abundance of potential benefits and use cases in the educational domain. In order for this potential to be realized, QG systems must be designed and validated with pedagogical needs in mind. However, little research has assessed or designed QG approaches with the input of real teachers or students. This paper applies a large language model-based QG approach where questions are generated with learning goals derived from Bloom's taxonomy. The automatically generated questions are used in multiple experiments designed to assess how teachers use them in practice. The results demonstrate that teachers prefer to write quizzes with automatically generated questions, and that such quizzes have no loss in quality compared to handwritten versions. Further, several metrics indicate that automatically generated questions can even improve the quality of the quizzes created, showing the promise for large scale use of QG in the classroom setting.

How Teachers Can Use Large Language Models and Bloom’s Taxonomy to Create Educational Quizzes | VIDEO

Large language models like ChatGPT can generate human-like code, posing challenges for programming education as students may be tempted to misuse them on assignments. However, there are currently no robust detectors designed specifically to identify AI-generated code. This is an issue that needs to be addressed to maintain academic integrity while allowing proper utilization of language models. Previous work has explored different approaches to detect AI-generated text, including watermarks, feature analysis, and fine-tuning language models. In this paper, we address the challenge of determining whether a student's code assignment was generated by a language model. First, our proposed method identifies AI-generated code by leveraging targeted masking perturbation paired with comperhesive scoring. Rather than applying a random mask, areas of the code with higher perplexity are more intensely masked. Second, we utilize a fine-tuned CodeBERT to fill in the masked portions, producing subtle modified samples. Then, we integrate the overall perplexity, variation of code line perplexity, and burstiness into a unified score. In this scoring scheme, a higher rank for the original code suggests it's more likely to be AI-generated. This approach stems from the observation that AI-generated codes typically have lower perplexity. Therefore, perturbations often exert minimal influence on them. Conversely, sections of human-composed codes that the model struggles to understand can see their perplexity reduced by such perturbations. Our method outperforms current open-source and commercial text detectors. Specifically, it improves detection of code submissions generated by OpenAI's text-davinci-003, raising average AUC from 0.56 (GPTZero baseline) to 0.87 for our detector.

Detecting AI-Generated Code Assignments Using Perplexity of Large Language Models | VIDEO

Training reinforcement learning (RL) agents to achieve desired goals while also acting morally is a challenging problem. Transformer-based language models (LMs) have shown some promise in moral awareness, but their use in different contexts is problematic because of the complexity and implicitness of human morality. In this paper, we build on text-based games, which are challenging environments for current RL agents, and propose the HuMAL (Human-guided Morality Awareness Learning) algorithm, which adaptively learns personal values through human-agent collaboration with minimal manual feedback. We evaluate HuMAL on the Jiminy Cricket benchmark, a set of text-based games with various scenes and dense morality annotations, using both simulated and actual human feedback. The experimental results demonstrate that with a small amount of human feedback, HuMAL can improve task performance and reduce immoral behavior in a variety of games and is adaptable to different personal values.

Human-Guided Moral Decision Making in Text-Based Games

Dataset sanitization is a widely adopted proactive defense against poisoning-based backdoor attacks, aimed at filtering out and removing poisoned samples from training datasets. However, existing methods have shown limited efficacy in countering the ever-evolving trigger functions, and often leading to considerable degradation of benign accuracy. In this paper, we propose DataElixir, a novel sanitization approach tailored to purify poisoned datasets. We leverage diffusion models to eliminate trigger features and restore benign features, thereby turning the poisoned samples into benign ones. Specifically, with multiple iterations of the forward and reverse process, we extract intermediary images and their predicted labels for each sample in the original dataset. Then, we identify anomalous samples in terms of the presence of label transition of the intermediary images, detect the target label by quantifying distribution discrepancy, select their purified images considering pixel and feature distance, and determine their ground-truth labels by training a benign model. Experiments conducted on 9 popular attacks demonstrates that DataElixir effectively mitigates various complex attacks while exerting minimal impact on benign accuracy, surpassing the performance of baseline defense methods.

DataElixir: Purifying Poisoned Dataset to Mitigate Backdoor Attacks via Diffusion Models

Recently, serious concerns have been raised about the privacy issues related to training datasets in machine learning algorithms when including personal data. Various regulations in different countries, including the GDPR grant individuals to have personal data erased, known as ‘the right to be forgotten’ or ‘the right to erasure’. However, there has been less research on effectively and practically deleting the requested personal data from the training set while not jeopardizing the overall machine learning performance. In this work, we propose a fast and novel machine unlearning paradigm at the layer level called layer attack unlearning, which is highly accurate and fast compared to existing machine unlearning algorithms. We introduce the Partial-PGD algorithm to locate the samples to forget efficiently. In addition, we only use the last layer of the model inspired by the Forward-Forward algorithm for unlearning process. Lastly, we use Knowledge Distillation (KD) to reliably learn the decision boundaries from the teacher using soft label information to improve accuracy performance. We conducted extensive experiments with SOTA machine unlearning models and demonstrated the effectiveness of our approach for accuracy and end-to-end unlearning performance.

Layer Attack Unlearning: Fast and Accurate Machine Unlearning via Layer Level Attack and Knowledge Distillation

Rescaling the backpropagated gradient of contrastive loss has made significant progress in descriptor learning. However, current gradient modulation strategies have no regard for the varying distribution of global gradients, so they would suffer from changes in training phases or datasets. In this paper, we propose a dynamic gradient modulation, named SDGMNet, for contrastive local descriptor learning. The core of our method is formulating modulation functions with dynamically estimated statistical characteristics. Firstly, we introduce angle for distance measure after deep analysis on backpropagation of pair-wise loss. On this basis, auto-focus modulation is employed to moderate the impact of statistically uncommon individual pairs in stochastic gradient descent optimization; probabilistic margin cuts off the gradients of proportional triplets that have achieved enough optimization; power adjustment balances the total weights of negative pairs and positive pairs. Extensive experiments demonstrate that our novel descriptor surpasses previous state-of-the-art methods in several tasks including patch verification, retrieval, pose estimation, and 3D reconstruction.

SDGMNet: Statistic-Based Dynamic Gradient Modulation for Local Descriptor Learning

The classical approach to AI was to design systems that were rational at run-time: they had explicit representations of beliefs, goals, and plans and ran inference algorithms, online, to select actions. The rational approach was criticized (by the behaviorists) and modified (by the probabilists) but persisted in some form. Now the overwhelming success of the connectionist approach in so many areas presents evidence that the rational view may no longer have a role to play in AI. I will examine this question from several perspectives, including whether the rationality is present at design-time and/or at run-time, and whether systems with run-time rationality might be useful from the perspectives of computational efficiency, cognitive modeling and safety. I will present some current research focused on understanding the roles of learning in runtime-rational systems with the ultimate aim of constructing general-purpose human-level intelligent robots.

The Role of Rationality in Modern AI

At AAAI 2024, we honour the legacy of Doug Lenat, a pioneering figure in artificial intelligence who founded Cycorp and profoundly impacted AI research by scaling both the extent and the ambition of logic-based and common-sense reasoning. This memorial and retrospective session brings together a distinguished panel of speakers close to Doug and the Cyc project, to reflect on Doug’s contributions to AI, particularly through his work on the Cyc project. Panelists, including Blake Shepard, Francesca Rossi, Gary Marcus, and Michael Witbrock, will discuss Lenat’s vision for AI, his groundbreaking approaches to knowledge representation and reasoning, and the enduring influence of his ideas on current and future AI research. We’ll explore how Lenat’s work on Cyc laid foundational principles for building intelligent systems and his advocacy for a comprehensive, inferentially powerful, AI knowledge base. Attendees will gain insights into Lenat’s impact on AI’s past, present, and its trajectory towards more sophisticated and human-like reasoning capabilities.

Doug Lenat, CYC, and future directions in reasoning and knowledge representation

Generative AI models such as ChatGPT and Stable Diffusion have greatly accelerated the development and adoption of AI applications. However, one of the key challenges today is AI hallucination – generative models tend to confabulate plausible even authoritative sounding responses that are incorrect and non-factual, leading to downstream harm if not mitigated. Some findings showed that close to 30% of chatbot responses are hallucinations. Why do they hallucinate? And why popular methods such as RLHF and safety fine-tuning cannot completely mitigate this problem? In this talk, I will give an overview of the known causes of hallucination and methods to mitigate them, including in multimodal and multilingual scenarios. I will also talk about the remaining challenges and future directions in AI hallucination research.

Downloads

Next from AAAI 2024

Does Any AI-Based Activity Contribute to Develop AI Conception? A Case Study with Italian Fifth and Sixth Grade Classes

Similar lecture

On the Concept Trustworthiness in Concept Bottleneck Models

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES