This practical experience report explores Neural Machine Translation (NMT) models' capability to generate offensive security code from natural language (NL) descriptions, highlighting the significance of contextual understanding and its impact on model performance. Our study employs a dataset comprising real shellcodes to evaluate the models across various scenarios, including missing information, necessary context, and unnecessary context. The experiments are designed to assess the models' resilience against incomplete descriptions, their proficiency in leveraging context for enhanced accuracy, and their ability to discern irrelevant information. The findings reveal that the introduction of contextual data significantly improves performance. However, the benefits of additional context diminish beyond a certain point, indicating an optimal level of contextual information for model training. Moreover, the models demonstrate an ability to filter out unnecessary context, maintaining high levels of accuracy in the generation of offensive security code. This study paves the way for future research on optimizing context use in AI-driven code generation, particularly for applications requiring a high degree of technical precision such as the generation of offensive code.

Vertical federated learning is a natural and elegant approach to integrate multi-view data vertically partitioned across devices (clients) while preserving their privacies. Apart from the model training, existing methods requires the collaboration of all clients in the model inference. However, the model inference is probably maintained for service in a long time, while the collaboration, especially when the clients belong to different organizations, is unpredictable in real-world scenarios, such as concellation of contract, network unavailablity, etc., resulting in the failure of them. To address this issue, we, at the first attempt, propose a flexible Active-Passive Federated learning (APFed) framework. Specifically, the active client is the initiator of a learning task and responsible to build the complete model, while the passive clients only serve as assistants. Once the model built, the active client can make inference independently. In addition, we instance the APFed framework into two classification methods with employing the reconstruction loss and the contrastive loss on passive clients, respectively. Meanwhile, the two methods are tested in a set of experiments and achieves desired results, validating their effectiveness.

LLMs have shown remarkable capabilities, but precisely controlling their response behavior remains challenging. Existing activation steering methods alter LLM behavior indiscriminately, limiting their practical applicability in settings where selective responses are essential, such as content moderation or domain-specific assistants. In this paper, we propose Conditional Activation Steering (CAST), which analyzes LLM activation patterns during inference to selectively apply or withhold activation steering based on the input context. Our method is based on the observation that different categories of prompts activate distinct patterns in the model's hidden states. Using CAST, one can systematically control LLM behavior with rules like "if input is about hate speech or adult content, then refuse" or "if input is not about legal advice, then refuse." This allows for selective modification of responses to specific content while maintaining normal responses to other content, all without requiring weight optimization. We release an open-source implementation of our framework.

Chain-of-Thought (CoT) prompting reveals that large language models are capable of performing complex reasoning via intermediate steps. CoT prompting is primarily categorized into three approaches. The first approach utilizes straightforward prompts like ``Let's think step by step'' to generate a sequential thought process before yielding an answer. The second approach makes use of human-crafted, step-by-step demonstrations to guide the model's reasoning process. The third automates the generation of reasoned demonstrations with the 'Let's think step by step'.This approach sometimes leads to reasoning errors, highlighting the need to diversify demonstrations to mitigate its misleading effects. However, diverse demonstrations pose challenges for effective representations. In this work, we propose ECHO, a self-harmonized chain-of-thought prompting method. It consolidates diverse solution paths into a uniform and effective solution pattern.ECHO demonstrates the best overall performance across three reasoning domains.

Line Current Differential Relays (LCDRs) are high-speed relays progressively used to protect critical transmission lines. However, LCDRs are vulnerable to cyberattacks. Fault-Masking Attacks (FMAs) are stealthy cyberattacks performed by manipulating the remote measurements of the targeted LCDR to disguise faults on the protected line. Hence, they remain undetected by this LCDR. In this paper, we propose a two-module framework to detect FMAs. The first module is a Mismatch Index (MI) developed from the protected transmission line's equivalent physical model. The MI is triggered only if there is a significant mismatch in the LCDR's local and remote measurements while the LCDR itself is untriggered, which indicates an FMA. After the MI is triggered, the second module, a neural network-based classifier, promptly confirms that the triggering event is a physical fault that lies on the line protected by the LCDR before declaring the occurrence of an FMA. The proposed framework is tested using the IEEE 39-bus benchmark system. Our simulation results confirm that the proposed framework can accurately detect FMAs on LCDRs and is not affected by normal system disturbances, variations, or measurement noise. Our experimental results using OPAL-RT's real-time simulator confirm the proposed solution's real-time performance capability.

Language style is necessary for AI systems to understand and generate diverse human language accurately. However, previous text style transfer primarily focused on sentence-level data-driven approaches, limiting exploration of potential problems in large language models (LLMs) and the ability to meet complex application needs. To overcome these limitations, we introduce a novel task called Public-Speaking Style Transfer (PSST), which aims to simulate humans to transform passage-level, official texts into a public-speaking style. Grounded in the analysis of real-world data from a linguistic perspective, we decompose public-speaking style into key sub-styles to pose challenges and quantify the style modeling capability of LLMs. For such intricate text style transfer, we further propose a fine-grained evaluation framework to analyze the characteristics and identify the problems of stylized texts. Comprehensive experiments suggest that current LLMs struggle to generate public speaking texts that align with human preferences, primarily due to excessive stylization and loss of semantic information.

Programming languages possess rich semantic information such as data flow that is represented by graphs and not available from the surface form of source code. Recent code language models have scaled to billions of parameters, but model source code solely as text tokens while ignoring any other structural information. Conversely, models that do encode structural information of code make modifications to the Transformer architecture, limiting their scale and compatibility with pretrained LLMs. In this work, we take the best of both worlds with GALLa - Graph Aligned Large Language Model. GALLa utilizes graph neural networks and cross-modal alignment technologies to inject the structural information of code into LLMs as an auxiliary task during finetuning. This framework is both model-agnostic and task-agnostic, as it can be applied to any code LLM for any code downstream task, and requires the structural graph data only at training time from a corpus unrelated to the finetuning data, while incurring no cost at inference time over the baseline LLM. Experiments on five code tasks with four different baseline LLMs ranging in size from 350M to 8B validate the effectiveness of GALLa, demonstrating consistent improvement over the baseline, even for powerful models such as LLaMA3.

Knowledge Organization Systems (KOSs), such as term lists, thesauri, taxonomies, and ontologies, play a fundamental role in categorising, managing, and retrieving information. In the academic domain, KOSs are often adopted for representing research areas and their relationships, primarily aiming to classify research articles, academic courses, patents, books, scientific venues, domain experts, grants, software, experiment materials, and several other relevant products and agents. These structured representations of research areas, widely embraced by many academic fields, have proven effective in empowering AI-based systems to i) enhance retrievability of relevant documents, ii) enable advanced analytic solutions to quantify the impact of academic research, and iii) analyse and forecast research dynamics. This paper aims to present a comprehensive survey of the current KOS for academic disciplines. We analysed and compared 45 KOSs according to five main dimensions: scope, structure, curation, usage, and links to other KOSs. Our results reveal a very heterogeneous scenario in terms of scope, scale, quality, and usage, highlighting the need for more integrated solutions for representing research knowledge across academic fields. We conclude by discussing the main challenges and the most promising future directions.

Language models can largely benefit from efficient tokenization. However, they still mostly utilize the classical BPE algorithm, a simple and reliable method. This has been shown to cause such issues as under-trained tokens and sub-optimal compression that may affect the downstream performance. We introduce Picky BPE, a modified BPE algorithm that carries Figure 1: An example of a series of merges to produce a out vocabulary refinement during tokenizer token Kentucky. The pre-merge token frequencies are training. Our method improves vocabulary efficiency, demonstrated in corresponding circles. In the vanilla eliminates under-trained tokens, and BPE algorithm, entucky should also be stored in the does not compromise text compression. Our vocabulary, whereas it is redundant after the merge. In experiments show that our method does not this example, the IoS metric effectively captures the reduce the downstream performance, and in intermediate token, as IoS(entucky) T = 0.9.

This document focuses on the threats, especially near-term threats, that Artificial Intelligence (AI) brings to society. Most of the threats discussed here can result from any algorithmic process, not just AI; in addition, defining AI is notoriously difficult. For both reasons, it is important to think of "A+AI": Algorithms and Artificial Intelligence. In addition to the threats, this paper discusses countermeasures to them, and it includes a table showing which countermeasures are likely to mitigate which threats. Thoughtful governance could manage the risks without seriously impeding progress; in fact, chances are it would accelerate progress by reducing the social chaos that would otherwise be likely.