Concerns regarding Large Language Models (LLMs) to memorize and disclose private information, particularly Personally Identifiable Information (PII), become prominent within the community. Many efforts have been made to mitigate the privacy risks. However, the mechanism through which LLMs memorize PII remains poorly understood. To bridge this gap, we introduce a pioneering method for pinpointing PII-sensitive neurons (privacy neurons) within LLMs. Our method employs learnable binary weight masks to localize specific neurons that account for the memorization of PII in LLMs through adversarial training. Our investigations discover that PII is memorized by a small subset of neurons across all layers, which shows the property of PII specificity. Furthermore, we propose to validate the potential in PII risk mitigation by deactivating the localized privacy neurons. Both quantitative and qualitative experiments demonstrate the effectiveness of our neuron localization algorithm.

Accurate evaluation of financial question answering (QA) systems necessitates a comprehensive dataset encompassing diverse question types and contexts. However, current financial QA datasets lack scope diversity and question complexity. This work introduces FinTextQA, a novel dataset for long-form question answering (LFQA) in finance. FinTextQA comprises 1,262 high-quality, source-attributed QA pairs extracted and selected from finance textbooks and government agency websites.Moreover, we developed a Retrieval-Augmented Generation (RAG)-based LFQA system, comprising an embedder, retriever, reranker, and generator. A multi-faceted evaluation approach, including human ranking, automatic metrics, and GPT-4 scoring, was employed to benchmark the performance of different LFQA system configurations under heightened noisy conditions. The results indicate that: (1) Among all compared generators, Baichuan2-7B competes closely with GPT-3.5-turbo in accuracy score; (2) The most effective system configuration on our dataset involved setting the embedder, retriever, reranker, and generator as Ada2, Automated Merged Retrieval, Bge-Reranker-Base, and Baichuan2-7B, respectively; (3) models are less susceptible to noise after the length of contexts reaching a specific threshold.

Stance detection classifies stance relations (namely, Favor, Against, or Neither) between comments and targets. Pretrained language models (PLMs) are widely used to mine the stance relation to improve the performance of stance detection through pretrained knowledge. However, PLMs also embed ``bad'' pretrained knowledge concerning stance into the extracted stance relation semantics, resulting in pretrained stance bias. It is not trivial to measure pretrained stance bias due to its weak quantifiability. In this paper, we propose Relative Counterfactual Contrastive Learning (RCCL), in which pretrained stance bias is mitigated as relative stance bias instead of absolute stance bias to overtake the difficulty of measuring bias. Firstly, we present a new structural causal model for characterizing complicated relationships among context, PLMs and stance relations to locate pretrained stance bias. Then, based on masked language model prediction, we present a target-aware relative stance sample generation method for obtaining relative bias. Finally, we use contrastive learning based on counterfactual theory to mitigate pretrained stance bias and preserve context stance relation. Experiments show that the proposed method is superior to stance detection and debiasing baselines.

AI's capabilities are growing at tremendous speeds, and while that apparently warrants a ton of the United States' money for development, it doesn't seem to translate to a very obvious action: regulation. A bipartisan group of four senators, led by majority leader Chuck Schumer, have announced a legislative plan for AI that includes putting 32 billion towards research and development. "It's very hard to do regulations because AI is changing too quickly," Schumer said in an interview published by The New York Times. Yet, in March, the European Parliament approved wide-ranging legislation for regulating AI that manages the obligations of AI applications based on what risks and effects they could bring. The European Union said it hopes to "protect fundamental rights, democracy, the rule of law and environmental sustainability from high-risk AI, while boosting innovation and establishing Europe as a leader in the field." Schumer seems to disagree with finding that balance, instead stating in the interview that investment into AI research and development "is sort of the American way -- we are more entrepreneurial."

The advent of transformers with attention mechanisms and associated pre-trained models have revolutionized the field of Natural Language Processing (NLP). However, such models are resource-intensive due to highly complex architecture. This limits their application to resource-constrained environments. While choosing an appropriate NLP model, a major trade-off exists over choosing accuracy over efficiency and vice versa. This paper presents a commentary on the evolution of NLP and its applications with emphasis on their accuracy as-well-as efficiency. Following this, a survey of research contributions towards enhancing the efficiency of transformer-based models at various stages of model development along with hardware considerations has been conducted. The goal of this survey is to determine how current NLP techniques contribute towards a sustainable society and to establish a foundation for future research.

The advent of advanced AI underscores the urgent need for comprehensive safety evaluations, necessitating collaboration across communities (i.e., AI, software engineering, and governance). However, divergent practices and terminologies across these communities, combined with the complexity of AI systems-of which models are only a part-and environmental affordances (e.g., access to tools), obstruct effective communication and comprehensive evaluation. This paper proposes a framework for AI system evaluation comprising three components: 1) harmonised terminology to facilitate communication across communities involved in AI safety evaluation; 2) a taxonomy identifying essential elements for AI system evaluation; 3) a mapping between AI lifecycle, stakeholders, and requisite evaluations for accountable AI supply chain. This framework catalyses a deeper discourse on AI system evaluation beyond model-centric approaches.

Information Retrieval (IR) systems are crucial tools for users to access information, widely applied in scenarios like search engines, question answering, and recommendation systems. Traditional IR methods, based on similarity matching to return ranked lists of documents, have been reliable means of information acquisition, dominating the IR field for years. With the advancement of pre-trained language models, generative information retrieval (GenIR) has emerged as a novel paradigm, gaining increasing attention in recent years. Currently, research in GenIR can be categorized into two aspects: generative document retrieval (GR) and reliable response generation. GR leverages the generative model's parameters for memorizing documents, enabling retrieval by directly generating relevant document identifiers without explicit indexing. Reliable response generation, on the other hand, employs language models to directly generate the information users seek, breaking the limitations of traditional IR in terms of document granularity and relevance matching, offering more flexibility, efficiency, and creativity, thus better meeting practical needs. This paper aims to systematically review the latest research progress in GenIR. We will summarize the advancements in GR regarding model training, document identifier, incremental learning, downstream tasks adaptation, multi-modal GR and generative recommendation, as well as progress in reliable response generation in aspects of internal knowledge memorization, external knowledge augmentation, generating response with citations and personal information assistant. We also review the evaluation, challenges and future prospects in GenIR systems. This review aims to offer a comprehensive reference for researchers in the GenIR field, encouraging further development in this area.

Federated learning (FL) enables collaborative model training while preserving data privacy, making it suitable for decentralized human-centered AI applications. However, a significant research gap remains in ensuring fairness in these systems. Current fairness strategies in FL require knowledge of bias-creating/sensitive attributes, clashing with FL's privacy principles. Moreover, in human-centered datasets, sensitive attributes may remain latent. To tackle these challenges, we present a novel bias mitigation approach inspired by "Fairness without Demographics" in machine learning. The presented approach achieves fairness without needing knowledge of sensitive attributes by minimizing the top eigenvalue of the Hessian matrix during training, ensuring equitable loss landscapes across FL participants. Notably, we introduce a novel FL aggregation scheme that promotes participating models based on error rates and loss landscape curvature attributes, fostering fairness across the FL system. This work represents the first approach to attaining "Fairness without Demographics" in human-centered FL. Through comprehensive evaluation, our approach demonstrates effectiveness in balancing fairness and efficacy across various real-world applications, FL setups, and scenarios involving single and multiple bias-inducing factors, representing a significant advancement in human-centered FL.

We study the problem of automatically annotating relevant numerals (GAAP metrics) occurring in the financial documents with their corresponding XBRL tags. Different from prior works, we investigate the feasibility of solving this extreme classification problem using a generative paradigm through instruction tuning of Large Language Models (LLMs). To this end, we leverage metric metadata information to frame our target outputs while proposing a parameter efficient solution for the task using LoRA. We perform experiments on two recently released financial numeric labeling datasets. Our proposed model, FLAN-FinXC, achieves new state-of-the-art performances on both the datasets, outperforming several strong baselines. We explain the better scores of our proposed model by demonstrating its capability for zero-shot as well as the least frequently occurring tags. Also, even when we fail to predict the XBRL tags correctly, our generated output has substantial overlap with the ground-truth in majority of the cases.

The rapid advancement of AI technologies yields numerous future impacts on individuals and society. Policy-makers are therefore tasked to react quickly and establish policies that mitigate those impacts. However, anticipating the effectiveness of policies is a difficult task, as some impacts might only be observable in the future and respective policies might not be applicable to the future development of AI. In this work we develop a method for using large language models (LLMs) to evaluate the efficacy of a given piece of policy at mitigating specified negative impacts. We do so by using GPT-4 to generate scenarios both pre- and post-introduction of policy and translating these vivid stories into metrics based on human perceptions of impacts. We leverage an already established taxonomy of impacts of generative AI in the media environment to generate a set of scenario pairs both mitigated and non-mitigated by the transparency legislation of Article 50 of the EU AI Act. We then run a user study (n=234) to evaluate these scenarios across four risk-assessment dimensions: severity, plausibility, magnitude, and specificity to vulnerable populations. We find that this transparency legislation is perceived to be effective at mitigating harms in areas such as labor and well-being, but largely ineffective in areas such as social cohesion and security. Through this case study on generative AI harms we demonstrate the efficacy of our method as a tool to iterate on the effectiveness of policy on mitigating various negative impacts. We expect this method to be useful to researchers or other stakeholders who want to brainstorm the potential utility of different pieces of policy or other mitigation strategies.