Large language models (LLMs) are vulnerable to adversarial attacks that can bypass their safety guardrails. In many domains, adversarial training has proven to be one of the most promising methods to reliably improve robustness against such attacks. Yet, in the context of LLMs, current methods for adversarial training are hindered by the high computational costs required to perform discrete adversarial attacks at each training iteration. We address this problem by instead calculating adversarial attacks in the continuous embedding space of the LLM, which is orders of magnitudes more efficient. We propose a fast adversarial training algorithm (C-AdvUL) composed of two losses: the first makes the model robust on continuous embedding attacks computed on an adversarial behaviour dataset; the second ensures the usefulness of the final model by fine-tuning on utility data. Moreover, we introduce C-AdvIPO, an adversarial variant of IPO that does not require utility data for adversarially robust alignment. Our empirical evaluation on four models from different families (Gemma, Phi3, Mistral, Zephyr) and at different scales (2B, 3.8B, 7B) shows that both algorithms substantially enhance LLM robustness against discrete attacks (GCG, AutoDAN, PAIR), while maintaining utility. Our results demonstrate that robustness to continuous perturbations can extrapolate to discrete threat models. Thereby, we present a path toward scalable adversarial training algorithms for robustly aligning LLMs.

NL2VIS (natural language to visualization) is a promising and recent research area that involves interpreting natural language queries and translating them into visualizations that accurately represent the underlying data. As we navigate the era of big data, NL2VIS holds considerable application potential since it greatly facilitates data exploration by non-expert users. Following the increasingly widespread usage of generative AI in NL2VIS applications, in this paper we present V-RECS, the first LLM-based Visual Recommender augmented with explanations(E), captioning(C), and suggestions(S) for further data exploration. V-RECS' visualization narratives facilitate both response verification and data exploration by non-expert users. Furthermore, our proposed solution mitigates computational, controllability, and cost issues associated with using powerful LLMs by leveraging a methodology to effectively fine-tune small models. To generate insightful visualization narratives, we use Chain-of-Thoughts (CoT), a prompt engineering technique to help LLM identify and generate the logical steps to produce a correct answer. Since CoT is reported to perform poorly with small LLMs, we adopted a strategy in which a large LLM (GPT-4), acting as a Teacher, generates CoT-based instructions to fine-tune a small model, Llama-2-7B, which plays the role of a Student. Extensive experiments-based on a framework for the quantitative evaluation of AI-based visualizations and on manual assessment by a group of participants-show that V-RECS achieves performance scores comparable to GPT-4, at a much lower cost. The efficacy of the V-RECS teacher-student paradigm is also demonstrated by the fact that the un-tuned Llama fails to perform the task in the vast majority of test cases. We release V-RECS for the visualization community to assist visualization designers throughout the entire visualization generation process.

In this paper, we present an end-to-end multi-source Entity Matching problem, which we call entity group matching, where the goal is to assign to the same group, records originating from multiple data sources but representing the same real-world entity. We focus on the effects of transitively matched records, i.e. the records connected by paths in the graph G = (V,E) whose nodes and edges represent the records and whether they are a match or not. We present a real-world instance of this problem, where the challenge is to match records of companies and financial securities originating from different data providers. We also introduce two new multi-source benchmark datasets that present similar matching challenges as real-world records. A distinctive characteristic of these records is that they are regularly updated following real-world events, but updates are not applied uniformly across data sources. This phenomenon makes the matching of certain groups of records only possible through the use of transitive information. In our experiments, we illustrate how considering transitively matched records is challenging since a limited amount of false positive pairwise match predictions can throw off the group assignment of large quantities of records. Thus, we propose GraLMatch, a method that can partially detect and remove false positive pairwise predictions through graph-based properties. Finally, we showcase how fine-tuning a Transformer-based model (DistilBERT) on a reduced number of labeled samples yields a better final entity group matching than training on more samples and/or incorporating fine-tuning optimizations, illustrating how precision becomes the deciding factor in the entity group matching of large volumes of records.

Trust plays a pivotal role in Lecturer-Student-Collaboration, encompassing teaching and research aspects. The advent of Large Language Models (LLMs) in platforms like Open AI's ChatGPT, coupled with their cost-effectiveness and high-quality results, has led to their rapid adoption among university students. However, discerning genuine student input from LLM-generated output poses a challenge for lecturers. This dilemma jeopardizes the trust relationship between lecturers and students, potentially impacting university downstream activities, particularly collaborative research initiatives. Despite attempts to establish guidelines for student LLM use, a clear framework mutually beneficial for lecturers and students in higher education remains elusive. This study addresses the research question: How does the use of LLMs by students impact Informational and Procedural Justice, influencing Team Trust and Expected Team Performance? Methodically, we applied a quantitative construct-based survey, evaluated using techniques of Structural Equation Modelling (PLS- SEM) to examine potential relationships among these constructs. Our findings based on 23 valid respondents from Ndejje University indicate that lecturers are less concerned about the fairness of LLM use per se but are more focused on the transparency of student utilization, which significantly influences Team Trust positively. This research contributes to the global discourse on integrating and regulating LLMs and subsequent models in education. We propose that guidelines should support LLM use while enforcing transparency in Lecturer-Student- Collaboration to foster Team Trust and Performance. The study contributes valuable insights for shaping policies enabling ethical and transparent LLMs usage in education to ensure effectiveness of collaborative learning environments.

Social scientists and the general public often analyze contemporary events by drawing parallels with the past, a process complicated by the vast, noisy, and unstructured nature of historical texts. For example, hundreds of millions of page scans from historical newspapers have been noisily transcribed. Traditional sparse methods for searching for relevant material in these vast corpora, e.g., with keywords, can be brittle given complex vocabularies and OCR noise. This study introduces News Deja Vu, a novel semantic search tool that leverages transformer large language models and a bi-encoder approach to identify historical news articles that are most similar to modern news queries. News Deja Vu first recognizes and masks entities, in order to focus on broader parallels rather than the specific named entities being discussed. Then, a contrastively trained, lightweight bi-encoder retrieves historical articles that are most similar semantically to a modern query, illustrating how phenomena that might seem unique to the present have varied historical precedents. Aimed at social scientists, the user-friendly News Deja Vu package is designed to be accessible for those who lack extensive familiarity with deep learning. It works with large text datasets, and we show how it can be deployed to a massive scale corpus of historical, open-source news articles. While human expertise remains important for drawing deeper insights, News Deja Vu provides a powerful tool for exploring parallels in how people have perceived past and present.

As large language models (LLMs) increasingly permeate daily lives, there is a growing demand for real-time interactions that mirror human conversations. Traditional turn-based chat systems driven by LLMs prevent users from verbally interacting with the system while it is generating responses. To overcome these limitations, we adapt existing LLMs to \textit{duplex models} so that these LLMs can listen for users while generating output and dynamically adjust themselves to provide users with instant feedback. % such as in response to interruptions. Specifically, we divide the queries and responses of conversations into several time slices and then adopt a time-division-multiplexing (TDM) encoding-decoding strategy to pseudo-simultaneously process these slices. Furthermore, to make LLMs proficient enough to handle real-time conversations, we build a fine-tuning dataset consisting of alternating time slices of queries and responses as well as covering typical feedback types in instantaneous interactions. Our experiments show that although the queries and responses of conversations are segmented into incomplete slices for processing, LLMs can preserve their original performance on standard benchmarks with a few fine-tuning steps on our dataset. Automatic and human evaluation indicate that duplex models make user-AI interactions more natural and human-like, and greatly improve user satisfaction compared to vanilla LLMs. Our duplex model and dataset will be released.

Recent advancements in Large Language Models (LLMs) have significantly increased their use in various real-world applications, including information retrieval and coding assistance [1]. Notably, the dramatic expansion of context window sizes in models like GPT-4 [2], Claude 3 [3], and Gemini-1.5 [4] has broadened the potential applications of these models. To evaluate the retrieval capabilities of these LLMs within large context windows, a series of benchmarks known as Needle-in-a-Haystack (NIAH) [5] has been developed. The NIAH benchmarks [5] typically involve prompting an LLM to retrieve contextual information based on a clue (e.g., needle) hidden within a large document (e.g., background). These benchmarks have been effective in evaluating LLMs' ability to retrieve information from large text data such as in text-summarization, and legal and medical domains [6, 7, 8]. NIAH represents important use-cases finding precedent case law in the legal domain [7] and information retrieval from lengthy electronic health records in the medical domain [8]. Verifying the "faithfulness" of long-text-summarization has also been shown as an important NIAH task for the FABLES dataset [6]. Generating code and programs following provided specifications or requirements is a long-standing challenge in computer science called program synthesis [9].

Traditional manufacturing faces challenges adapting to dynamic environments and quickly responding to manufacturing changes. The use of multi-agent systems has improved adaptability and coordination but requires further advancements in rapid human instruction comprehension, operational adaptability, and coordination through natural language integration. Large language models like GPT-3.5 and GPT-4 enhance multi-agent manufacturing systems by enabling agents to communicate in natural language and interpret human instructions for decision-making. This research introduces a novel framework where large language models enhance the capabilities of agents in manufacturing, making them more adaptable, and capable of processing context-specific instructions. A case study demonstrates the practical application of this framework, showing how agents can effectively communicate, understand tasks, and execute manufacturing processes, including precise G-code allocation among agents. The findings highlight the importance of continuous large language model integration into multi-agent manufacturing systems and the development of sophisticated agent communication protocols for a more flexible manufacturing system.

The increasing size of language models necessitates a thorough analysis across multiple dimensions to assess trade-offs among crucial hardware metrics such as latency, energy consumption, GPU memory usage, and performance. Identifying optimal model configurations under specific hardware constraints is becoming essential but remains challenging due to the computational load of exhaustive training and evaluation on multiple devices. To address this, we introduce HW-GPT-Bench, a hardware-aware benchmark that utilizes surrogate predictions to approximate various hardware metrics across 13 devices of architectures in the GPT-2 family, with architectures containing up to 774M parameters. Our surrogates, via calibrated predictions and reliable uncertainty estimates, faithfully model the heteroscedastic noise inherent in the energy and latency measurements. To estimate perplexity, we employ weight-sharing techniques from Neural Architecture Search (NAS), inheriting pretrained weights from the largest GPT-2 model. Finally, we demonstrate the utility of HW-GPT-Bench by simulating optimization trajectories of various multi-objective optimization algorithms in just a few seconds.

Understanding the extent to which Chain-of-Thought (CoT) generations align with a large language model's (LLM) internal computations is critical for deciding whether to trust an LLM's output. As a proxy for CoT faithfulness, Lanham et al. (2023) propose a metric that measures a model's dependence on its CoT for producing an answer. Within a single family of proprietary models, they find that LLMs exhibit a scaling-then-inverse-scaling relationship between model size and their measure of faithfulness, and that a 13 billion parameter model exhibits increased faithfulness compared to models ranging from 810 million to 175 billion parameters in size. We evaluate whether these results generalize as a property of all LLMs. We replicate the experimental setup in their section focused on scaling experiments with three different families of models and, under specific conditions, successfully reproduce the scaling trends for CoT faithfulness they report. However, after normalizing the metric to account for a model's bias toward certain answer choices, unfaithfulness drops significantly for smaller less-capable models. This normalized faithfulness metric is also strongly correlated ($R^2$=0.74) with accuracy, raising doubts about its validity for evaluating faithfulness.