Intelligent agents powered by large language models (LLMs) have demonstrated substantial promise in autonomously conducting experiments and facilitating scientific discoveries across various disciplines. While their capabilities are promising, they also introduce novel vulnerabilities that demand careful consideration for safety. However, there exists a notable gap in the literature, as there has been no comprehensive exploration of these vulnerabilities. This position paper fills this gap by conducting a thorough examination of vulnerabilities in LLM-based agents within scientific domains, shedding light on potential risks associated with their misuse and emphasizing the need for safety measures. We begin by providing a comprehensive overview of the potential risks inherent to scientific LLM agents, taking into account user intent, the specific scientific domain, and their potential impact on the external environment. Then, we delve into the origins of these vulnerabilities and provide a scoping review of the limited existing works. Based on our analysis, we propose a triadic framework involving human regulation, agent alignment, and an understanding of environmental feedback (agent regulation) to mitigate these identified risks. Furthermore, we highlight the limitations and challenges associated with safeguarding scientific agents and advocate for the development of improved models, robust benchmarks, and comprehensive regulations to address these issues effectively.

Large Language Models (LLMs) have demonstrated a number of human-like abilities, however the empathic understanding and emotional state of LLMs is yet to be aligned to that of humans. In this work, we investigate how the emotional state of generative LLM agents evolves as they perceive new events, introducing a novel architecture in which new experiences are compared to past memories. Through this comparison, the agent gains the ability to understand new experiences in context, which according to the appraisal theory of emotion is vital in emotion creation. First, the agent perceives new experiences as time series text data. After perceiving each new input, the agent generates a summary of past relevant memories, referred to as the norm, and compares the new experience to this norm. Through this comparison we can analyse how the agent reacts to the new experience in context. The PANAS, a test of affect, is administered to the agent, capturing the emotional state of the agent after the perception of the new event. Finally, the new experience is then added to the agents memory to be used in the creation of future norms. By creating multiple experiences in natural language from emotionally charged situations, we test the proposed architecture on a wide range of scenarios. The mixed results suggests that introducing context can occasionally improve the emotional alignment of the agent, but further study and comparison with human evaluators is necessary. We hope that this paper is another step towards the alignment of generative agents.

Developing a generalist agent is a longstanding objective in artificial intelligence. Previous efforts utilizing extensive offline datasets from various tasks demonstrate remarkable performance in multitasking scenarios within Reinforcement Learning. However, these works encounter challenges in extending their capabilities to new tasks. Recent approaches integrate textual guidance or visual trajectory into decision networks to provide task-specific contextual cues, representing a promising direction. However, it is observed that relying solely on textual guidance or visual trajectory is insufficient for accurately conveying the contextual information of tasks. This paper explores enhanced forms of task guidance for agents, enabling them to comprehend gameplay instructions, thereby facilitating a "read-to-play" capability. Drawing inspiration from the success of multimodal instruction tuning in visual tasks, we treat the visual-based RL task as a long-horizon vision task and construct a set of multimodal game instructions to incorporate instruction tuning into a decision transformer. Experimental results demonstrate that incorporating multimodal game instructions significantly enhances the decision transformer's multitasking and generalization capabilities.

This study consists of a novel approach toward the analysis of court judgments spanning five countries, including the United States, the United Kingdom, Rwanda, Sweden and Hong Kong. This study also explores the intersection of the latest advancements in artificial intelligence (AI) and legal analysis, emphasizing the role of AI (specifically generative AI) in identifying human biases and facilitating automated, valid, and coherent multisided argumentation of court judgments with the goal of ensuring consistent application of laws in and across various jurisdictions. By incorporating Advanced Language Models (ALMs) and a newly introduced human-AI collaborative framework, this paper seeks to analyze Grounded Theory-based research design with Advanced Language Models (ALMs) in the practice of law. SHIRLEY is the name of the AI-based application (built on top of OpenAI's GPT technology), focusing on detecting logical inconsistencies and biases across various legal decisions. SHIRLEY analysis is aggregated and is accompanied by a comparison-oriented AI-based application called SAM (also an ALM) to identify relative deviations in SHIRLEY bias detections. Further, a CRITIC is generated within semi-autonomous arbitration process via the ALM, SARA. A novel approach is introduced in the utilization of an AI arbitrator to critically evaluate biases and qualitative-in-nature nuances identified by the aforementioned AI applications (SAM in concert with SHIRLEY), based on the Hague Rules on Business and Human Rights Arbitration. This Semi-Automated Arbitration Process (SAAP) aims to uphold the integrity and fairness of legal judgments by ensuring a nuanced debate-resultant "understanding" through a hybrid system of AI and human-based collaborative analysis.

Humans have a tendency to see 'human'-like qualities in objects around them. We name our cars, and talk to pets and even household appliances, as if they could understand us as other humans do. This behavior, called anthropomorphism, is also seeing traction in Machine Learning (ML), where human-like intelligence is claimed to be perceived in Large Language Models (LLMs). In this position paper, considering professional incentives, human biases, and general methodological setups, we discuss how the current search for Artificial General Intelligence (AGI) is a perfect storm for over-attributing human-like qualities to LLMs. In several experiments, we demonstrate that the discovery of human-interpretable patterns in latent spaces should not be a surprising outcome. Also in consideration of common AI portrayal in the media, we call for the academic community to exercise extra caution, and to be extra aware of principles of academic integrity, in interpreting and communicating about AI research outcomes.

In this work we study the enhancement of Low Rank Adaptation (LoRA) fine-tuning procedure by introducing a Riemannian preconditioner in its optimization step. Specifically, we introduce an $r\times r$ preconditioner in each gradient step where $r$ is the LoRA rank. This preconditioner requires a small change to existing optimizer code and creates virtually minuscule storage and runtime overhead. Our experimental results with both large language models and text-to-image diffusion models show that with our preconditioner, the convergence and reliability of SGD and AdamW can be significantly enhanced. Moreover, the training process becomes much more robust to hyperparameter choices such as learning rate. Theoretically, we show that fine-tuning a two-layer ReLU network in the convex paramaterization with our preconditioner has convergence rate independent of condition number of the data matrix. This new Riemannian preconditioner, previously explored in classic low-rank matrix recovery, is introduced to deep learning tasks for the first time in our work. We release our code at https://github.com/pilancilab/Riemannian_Preconditioned_LoRA.

Large language models (LLMs) and foundation models have been recently touted as a game-changer for 6G systems. However, recent efforts on LLMs for wireless networks are limited to a direct application of existing language models that were designed for natural language processing (NLP) applications. To address this challenge and create wireless-centric foundation models, this paper presents a comprehensive vision on how to design universal foundation models that are tailored towards the deployment of artificial intelligence (AI)-native networks. Diverging from NLP-based foundation models, the proposed framework promotes the design of large multi-modal models (LMMs) fostered by three key capabilities: 1) processing of multi-modal sensing data, 2) grounding of physical symbol representations in real-world wireless systems using causal reasoning and retrieval-augmented generation (RAG), and 3) enabling instructibility from the wireless environment feedback to facilitate dynamic network adaptation thanks to logical and mathematical reasoning facilitated by neuro-symbolic AI. In essence, these properties enable the proposed LMM framework to build universal capabilities that cater to various cross-layer networking tasks and alignment of intents across different domains. Preliminary results from experimental evaluation demonstrate the efficacy of grounding using RAG in LMMs, and showcase the alignment of LMMs with wireless system designs. Furthermore, the enhanced rationale exhibited in the responses to mathematical questions by LMMs, compared to vanilla LLMs, demonstrates the logical and mathematical reasoning capabilities inherent in LMMs. Building on those results, we present a sequel of open questions and challenges for LMMs. We then conclude with a set of recommendations that ignite the path towards LMM-empowered AI-native systems.

For many people, anxiety, depression, and other social and mental factors can make composing text messages an active challenge. To remedy this problem, large language models (LLMs) may yet prove to be the perfect tool to assist users that would otherwise find texting difficult or stressful. However, despite rapid uptake in LLM usage, considerations for their assistive usage in text message composition have not been explored. A primary concern regarding LLM usage is that poor public sentiment regarding AI introduces the possibility that its usage may harm perceptions of AI-assisted text messages, making usage counter-productive. To (in)validate this possibility, we explore how the belief that a text message did or did not receive AI assistance in composition alters its perceived tone, clarity, and ability to convey intent. In this study, we survey the perceptions of 26 participants on 18 randomly labeled pre-composed text messages. In analyzing the participants' ratings of message tone, clarity, and ability to convey intent, we find that there is no statistically significant evidence that the belief that AI is utilized alters recipient perceptions. This provides hopeful evidence that LLM-based text message composition assistance can be implemented without the risk of counter-productive outcomes.

Large language models (LLMs) are not amenable to frequent re-training, due to high training costs arising from their massive scale. However, updates are necessary to endow LLMs with new skills and keep them up-to-date with rapidly evolving human knowledge. This paper surveys recent works on continual learning for LLMs. Due to the unique nature of LLMs, we catalog continue learning techniques in a novel multi-staged categorization scheme, involving continual pretraining, instruction tuning, and alignment. We contrast continual learning for LLMs with simpler adaptation methods used in smaller models, as well as with other enhancement strategies like retrieval-augmented generation and model editing. Moreover, informed by a discussion of benchmarks and evaluation, we identify several challenges and future work directions for this crucial task.

Tokenization is an understudied and often neglected component of modern LLMs. Most published works use a single tokenizer for all experiments, often borrowed from another model, without performing ablations or analysis to optimize tokenization. Moreover, the tokenizer is generally kept unchanged when fine-tuning a base model. In this paper, we show that the size, pre-tokenization regular expression, and training data of a tokenizer can significantly impact the model's generation speed, effective context size, memory usage, and downstream performance. We train specialized Byte-Pair Encoding code tokenizers, and conduct extensive ablations on the impact of tokenizer design on the performance of LLMs for code generation tasks such as HumanEval and MBPP, and provide recommendations for tokenizer hyper-parameters selection and switching the tokenizer in a pre-trained LLM. We perform our experiments on models trained from scratch and from pre-trained models, verifying their applicability to a wide range of use-cases. We find that when fine-tuning on more than 50 billion tokens, we can specialize the tokenizer of a pre-trained LLM to obtain large gains in generation speed and effective context size.