Preference optimization is a standard approach to fine-tuning large language models to align with human preferences. The quality, diversity, and quantity of the preference dataset are critical to the effectiveness of preference optimization. However, obtaining a large amount of high-quality and diverse preference annotations is difficult in many applications. This raises the question of how to use the limited annotation budget to create an effective preference dataset. To this end, we propose Annotation-Efficient Preference Optimization (AEPO). Instead of exhaustively annotating preference over all available response texts, AEPO selects a subset of responses that maximizes quality and diversity from the available responses, and then annotates preference over the selected ones. In this way, AEPO focuses the annotation budget on labeling preference over a smaller subset of responses with diversity and of high quality. We evaluate the performance of Direct Preference Optimization (DPO) using AEPO and show that it outperforms models trained using a standard DPO with the same annotation budget.

Modern code review is a process in which an incremental code contribution made by a code author is reviewed by one or more peers before it is committed to the version control system. An important element of modern code review is verifying that code contributions adhere to best practices. While some of these best practices can be automatically verified, verifying others is commonly left to human reviewers. This paper reports on the development, deployment, and evaluation of AutoCommenter, a system backed by a large language model that automatically learns and enforces coding best practices. We implemented AutoCommenter for four programming languages (C++, Java, Python, and Go) and evaluated its performance and adoption in a large industrial setting. Our evaluation shows that an end-to-end system for learning and enforcing coding best practices is feasible and has a positive impact on the developer workflow. Additionally, this paper reports on the challenges associated with deploying such a system to tens of thousands of developers and the corresponding lessons learned.

Recent developments in Parameter-Efficient Fine-Tuning (PEFT) methods for pretrained deep neural networks have captured widespread interest. In this work, we study the enhancement of current PEFT methods by incorporating the spectral information of pretrained weight matrices into the fine-tuning procedure. We investigate two spectral adaptation mechanisms, namely additive tuning and orthogonal rotation of the top singular vectors, both are done via first carrying out Singular Value Decomposition (SVD) of pretrained weights and then fine-tuning the top spectral space. We provide a theoretical analysis of spectral fine-tuning and show that our approach improves the rank capacity of low-rank adapters given a fixed trainable parameter budget. We show through extensive experiments that the proposed fine-tuning model enables better parameter efficiency and tuning performance as well as benefits multi-adapter fusion. The code will be open-sourced for reproducibility.

Prompting methods play a crucial role in enhancing the capabilities of pre-trained large language models (LLMs). We explore how contrastive prompting (CP) significantly improves the ability of large language models to perform complex reasoning. We demonstrate that LLMs are decent contrastive reasoners by simply adding "Let's give a correct and a wrong answer." before LLMs provide answers. Experiments on various large language models show that zero-shot contrastive prompting improves performance on a range of arithmetic, commonsense, and symbolic reasoning tasks without any hand-crafted few-shot examples, such as increasing the accuracy on GSM8K from 35.9% to 88.8% and AQUA-RAT from 41.3% to 62.2% with the state-of-the-art GPT-4 model. Our method not only surpasses zero-shot CoT and few-shot CoT in most arithmetic and commonsense reasoning tasks but also can seamlessly integrate with existing prompting methods, resulting in improved or comparable results when compared to state-of-the-art methods. Our code is available at https://github.com/yao8839836/cp

The Twelfth International Conference on Learning Representations (ICLR2024) took place from 7-11 May in Vienna. The program included workshops, contributed talks, affinity group events, and socials. There were also seven invited talks that covered a broad range of topics. In this post, we give a summary of the talk by Priya Donti. Priya's research focuses on machine learning for forecasting, optimization, and control in power grids. She is an Assistant Professor and the Silverman (1968) Family Career Development Professor at MIT.

A major challenge facing the world is the provision of equitable and universal access to quality education. Recent advances in generative AI (gen AI) have created excitement about the potential of new technologies to offer a personal tutor for every learner and a teaching assistant for every teacher. The full extent of this dream, however, has not yet materialised. We argue that this is primarily due to the difficulties with verbalising pedagogical intuitions into gen AI prompts and the lack of good evaluation practices, reinforced by the challenges in defining excellent pedagogy. Here we present our work collaborating with learners and educators to translate high level principles from learning science into a pragmatic set of seven diverse educational benchmarks, spanning quantitative, qualitative, automatic and human evaluations; and to develop a new set of fine-tuning datasets to improve the pedagogical capabilities of Gemini, introducing LearnLM-Tutor. Our evaluations show that LearnLM-Tutor is consistently preferred over a prompt tuned Gemini by educators and learners on a number of pedagogical dimensions. We hope that this work can serve as a first step towards developing a comprehensive educational evaluation framework, and that this can enable rapid progress within the AI and EdTech communities towards maximising the positive impact of gen AI in education.

The rise of Generative AI (GenAI) brings about transformative potential across sectors, but its dual-use nature also amplifies risks. Governments globally are grappling with the challenge of regulating GenAI, balancing innovation against safety. China, the United States (US), and the European Union (EU) are at the forefront with initiatives like the Management of Algorithmic Recommendations, the Executive Order, and the AI Act, respectively. However, the rapid evolution of GenAI capabilities often outpaces the development of comprehensive safety measures, creating a gap between regulatory needs and technical advancements. A workshop co-organized by Google, University of Wisconsin, Madison (UW-Madison), and Stanford University aimed to bridge this gap between GenAI policy and technology. The diverse stakeholders of the GenAI space -- from the public and governments to academia and industry -- make any safety measures under consideration more complex, as both technical feasibility and regulatory guidance must be realized. This paper summarizes the discussions during the workshop which addressed questions, such as: How regulation can be designed without hindering technological progress? How technology can evolve to meet regulatory standards? The interplay between legislation and technology is a very vast topic, and we don't claim that this paper is a comprehensive treatment on this topic. This paper is meant to capture findings based on the workshop, and hopefully, can guide discussion on this topic.

This paper provides a comprehensive review of the future of cybersecurity through Generative AI and Large Language Models (LLMs). We explore LLM applications across various domains, including hardware design security, intrusion detection, software engineering, design verification, cyber threat intelligence, malware detection, and phishing detection. We present an overview of LLM evolution and its current state, focusing on advancements in models such as GPT-4, GPT-3.5, Mixtral-8x7B, BERT, Falcon2, and LLaMA. Our analysis extends to LLM vulnerabilities, such as prompt injection, insecure output handling, data poisoning, DDoS attacks, and adversarial instructions. We delve into mitigation strategies to protect these models, providing a comprehensive look at potential attack scenarios and prevention techniques. Furthermore, we evaluate the performance of 42 LLM models in cybersecurity knowledge and hardware security, highlighting their strengths and weaknesses. We thoroughly evaluate cybersecurity datasets for LLM training and testing, covering the lifecycle from data creation to usage and identifying gaps for future research. In addition, we review new strategies for leveraging LLMs, including techniques like Half-Quadratic Quantization (HQQ), Reinforcement Learning with Human Feedback (RLHF), Direct Preference Optimization (DPO), Quantized Low-Rank Adapters (QLoRA), and Retrieval-Augmented Generation (RAG). These insights aim to enhance real-time cybersecurity defenses and improve the sophistication of LLM applications in threat detection and response. Our paper provides a foundational understanding and strategic direction for integrating LLMs into future cybersecurity frameworks, emphasizing innovation and robust model deployment to safeguard against evolving cyber threats.

In the current era of digital communication and widespread use of social media, it is crucial to develop an understanding of persuasive techniques employed in written text. This knowledge is essential for effectively discerning accurate information and making informed decisions. To address this need, this paper presents a comprehensive empirical study focused on identifying persuasive techniques in Arabic social media content. To achieve this objective, we utilize Pre-trained Language Models (PLMs) and leverage the ArAlEval dataset, which encompasses two tasks: binary classification to determine the presence or absence of persuasion techniques, and multi-label classification to identify the specific types of techniques employed in the text. Our study explores three different learning approaches by harnessing the power of PLMs: feature extraction, fine-tuning, and prompt engineering techniques. Through extensive experimentation, we find that the fine-tuning approach yields the highest results on the aforementioned dataset, achieving an f1-micro score of 0.865 and an f1-weighted score of 0.861. Furthermore, our analysis sheds light on an interesting finding. While the performance of the GPT model is relatively lower compared to the other approaches, we have observed that by employing few-shot learning techniques, we can enhance its results by up to 20\%. This offers promising directions for future research and exploration in this topic\footnote{Upon Acceptance, the source code will be released on GitHub.}.

Graph Neural Networks (GNNs) have garnered widespread attention for their potential to address the challenges posed by graph representation learning, which face complex graph-structured data across various domains. However, due to the inherent complexity and interconnectedness of graphs, accurately annotating graph data for training GNNs is extremely challenging. To address this issue, we have introduced Partial Label Learning (PLL) into graph representation learning. PLL is a critical weakly supervised learning problem where each training instance is associated with a set of candidate labels, including the ground-truth label and the additional interfering labels. PLL allows annotators to make errors, which reduces the difficulty of data labeling. Subsequently, we propose a novel graph representation learning method that enables GNN models to effectively learn discriminative information within the context of PLL. Our approach utilizes potential cause extraction to obtain graph data that holds causal relationships with the labels. By conducting auxiliary training based on the extracted graph data, our model can effectively eliminate the interfering information in the PLL scenario. We support the rationale behind our method with a series of theoretical analyses. Moreover, we conduct extensive evaluations and ablation studies on multiple datasets, demonstrating the superiority of our proposed method.