Layer-wise normalization (LN) is an essential component of virtually all transformer-based large language models. While its effects on training stability are well documented, its role at inference time is poorly understood. Additionally, LN layers hinder mechanistic interpretability by introducing additional nonlinearities and increasing the interconnectedness of individual model components. Here, we show that all LN layers can be removed from every GPT-2 model with only a small increase in validation loss (e.g. +0.03 cross-entropy loss for GPT-2 XL). Thus, LN cannot play a substantial role in language modeling. We find that the amount of fine-tuning data needed for LN removal grows sublinearly with model parameters, suggesting scaling to larger models is feasible. We release a suite of LN-free GPT-2 models on Hugging Face. Furthermore, we test interpretability techniques on LN-free models. Direct logit attribution now gives the exact direct effect of individual components, while the accuracy of attribution patching does not significantly improve. We also confirm that GPT-2's "confidence neurons" are inactive in the LN-free models. Our work clarifies the role of LN layers in language modeling, showing that GPT-2-class models can function without LN layers. We hope that our LN-free analogs of the GPT-2 family of models will enable more precise interpretability research and improve our understanding of language models.

Software systems must comply with legal regulations, which is a resource-intensive task, particularly for small organizations and startups lacking dedicated legal expertise. Extracting metadata from regulations to elicit legal requirements for software is a critical step to ensure compliance. However, it is a cumbersome task due to the length and complex nature of legal text. Although prior work has pursued automated methods for extracting structural and semantic metadata from legal text, key limitations remain: they do not consider the interplay and interrelationships among attributes associated with these metadata types, and they rely on manual labeling or heuristic-driven machine learning, which does not generalize well to new documents. In this paper, we introduce an approach based on textual entailment and in-context learning for automatically generating a canonical representation of legal text, encodable and executable as Python code. Our representation is instantiated from a manually designed Python class structure that serves as a domain-specific metamodel, capturing both structural and semantic legal metadata and their interrelationships. This design choice reduces the need for large, manually labeled datasets and enhances applicability to unseen legislation. We evaluate our approach on 13 U.S. state data breach notification laws, demonstrating that our generated representations pass approximately 89.4% of test cases and achieve a precision and recall of 82.2 and 88.7, respectively.

Fraudulent activity in the financial industry costs billions annually. Detecting fraud, therefore, is an essential yet technically challenging task that requires carefully analyzing large volumes of data. While machine learning (ML) approaches seem like a viable solution, applying them successfully is not so easy due to two main challenges: (1) the sparsely labeled data, which makes the training of such approaches challenging (with inherent labeling costs), and (2) lack of explainability for the flagged items posed by the opacity of ML models, that is often required by business regulations. This article proposes SAGE-FIN, a semi-supervised graph neural network (GNN) based approach with Granger causal explanations for Financial Interaction Networks. SAGE-FIN learns to flag fraudulent items based on weakly labeled (or unlabelled) data points. To adhere to regulatory requirements, the flagged items are explained by highlighting related items in the network using Granger causality. We empirically validate the favorable performance of SAGE-FIN on a real-world dataset, Bipartite Edge-And-Node Attributed financial network (Elliptic++), with Granger-causal explanations for the identified fraudulent items without any prior assumption on the network structure.

Information retrieval is a cornerstone of modern knowledge acquisition, enabling billions of queries each day across diverse domains. However, traditional keyword-based search engines are increasingly inadequate for handling complex, multi-step information needs. Our position is that Large Language Models (LLMs), endowed with reasoning and agentic capabilities, are ushering in a new paradigm termed Agentic Deep Research. These systems transcend conventional information search techniques by tightly integrating autonomous reasoning, iterative retrieval, and information synthesis into a dynamic feedback loop. We trace the evolution from static web search to interactive, agent-based systems that plan, explore, and learn. We also introduce a test-time scaling law to formalize the impact of computational depth on reasoning and search. Supported by benchmark results and the rise of open-source implementations, we demonstrate that Agentic Deep Research not only significantly outperforms existing approaches, but is also poised to become the dominant paradigm for future information seeking. All the related resources, including industry products, research papers, benchmark datasets, and open-source implementations, are collected for the community in https://github.com/DavidZWZ/Awesome-Deep-Research.

Elon Musk's artificial intelligence company xAI has been granted a permit to run methane gas generators at its massive datacenter in Memphis, Tennessee. The county health department approved the permit for the 15 machines late on Wednesday, a move that has sparked outcry from the local community and environmental leaders, who say the generators pollute their neighborhoods. "Our local leaders are entrusted with protecting us from corporations violating on our right to clean air, but we are witnessing their failure to do so," said KeShaun Pearson, the director of the local environmental non-profit Memphis Community Against Pollution. To supplement the facility's heavy power usage, the company brought in dozens of portable methane gas generators. In January, xAI did apply for a permit for 15 generators – even though it had been running up to 35 generators on-site, according to photographs.

The Department of Justice unveiled charges against 300 defendants, alleging they misled patients into paying for, and sometimes receiving, medical care they did not need in a 14.6 billion healthcare fraud scheme. The Department of Justice's unveiling this week of sweeping charges against more than 300 defendants who allegedly defrauded Medicare and other taxpayer-funded programs came as part of the department's annual "takedown" event. The healthcare fraud takedowns have been a practice at the DOJ for more than a decade, but officials touted this one as the largest on record. It stood out not only for its size but also because it focused on transnational criminals and broached artificial intelligence. "This takedown represents the largest healthcare fraud takedown in American history," DOJ Criminal Division head Matthew Galeotti said.

The judiciary, as one of democracy's three pillars, is dealing with a rising amount of legal issues, needing careful use of judicial resources. This research presents a complex framework that leverages Data Science methodologies, notably Large Language Models (LLM) and Retrieval-Augmented Generation (RAG) techniques, to improve the efficiency of analyzing Calcutta High Court verdicts. Our framework focuses on two key aspects: first, the creation of a robust summarization mechanism that distills complex legal texts into concise and coherent summaries; and second, the development of an intelligent system for retrieving similar cases, which will assist legal professionals in research and decision making. By fine-tuning the Pegasus model using case head note summaries, we achieve significant improvements in the summarization of legal cases. Our two-step summarizing technique preserves crucial legal contexts, allowing for the production of a comprehensive vector database for RAG. The RAG-powered framework efficiently retrieves similar cases in response to user queries, offering thorough overviews and summaries. This technique not only improves legal research efficiency, but it also helps legal professionals and students easily acquire and grasp key legal information, benefiting the overall legal scenario.

Recent advancements in Large Language Models (LLMs) have significantly enhanced the ability to develop systems that comprehend customer requests and determine the necessary actions to fulfill them. In today's competitive market, delivering superior custome r service is crucial for attracting and retaining clients. Satisfied customers are more likely to become loyal, repeat buyers, and advocate for your brand, leading to increased revenue and market share (Strikingly, 2024) . In industries characterized by intense competition, implementing LLM - based services that effectively address customer needs and enhance satisfaction is becoming a key determinant of a company's growth and success. By leveraging LLMs, businesses can deliver more personalized, efficient, and scalable support, and thereby improve customer experience and foster loyalty (Iopex, 2024) .

Despite the recent strides in large language models, studies have underscored the existence of social biases within these systems. In this paper, we delve into the validation and comparison of the ethical biases of LLMs concerning globally discussed and potentially sensitive topics, hypothesizing that these biases may arise from language-specific distinctions. Introducing the Multilingual Sensitive Questions & Answers Dataset (MSQAD), we collected news articles from Human Rights Watch covering 17 topics, and generated socially sensitive questions along with corresponding responses in multiple languages. We scrutinized the biases of these responses across languages and topics, employing two statistical hypothesis tests. The results showed that the null hypotheses were rejected in most cases, indicating biases arising from cross-language differences. It demonstrates that ethical biases in responses are widespread across various languages, and notably, these biases were prevalent even among different LLMs. By making the proposed MSQAD openly available, we aim to facilitate future research endeavors focused on examining cross-language biases in LLMs and their variant models.

We propose a reinforcement learning-based approach to optimize conversational strategies for product recommendation across diverse industries. As organizations increasingly adopt intelligent agents to support sales and service operations, the effectiveness of a conversation hinges not only on what is recommended but how and when recommendations are delivered. We explore a methodology where agentic systems learn optimal dialogue policies through feedback-driven reinforcement learning. By mining aggregate behavioral patterns and conversion outcomes, our approach enables agents to refine talk tracks that drive higher engagement and product uptake, while adhering to contextual and regulatory constraints. We outline the conceptual framework, highlight key innovations, and discuss the implications for scalable, personalized recommendation in enterprise environments.