Prior research has demonstrated that language models can, to a limited extent, represent moral norms in a variety of cultural contexts. This research aims to replicate these findings and further explore their validity, concentrating on issues like 'homosexuality' and 'divorce'. This study evaluates the effectiveness of these models using information from two surveys, the WVS and the PEW, that encompass moral perspectives from over 40 countries. The results show that biases exist in both monolingual and multilingual models, and they typically fall short of accurately capturing the moral intricacies of diverse cultures. However, the BLOOM model shows the best performance, exhibiting some positive correlations, but still does not achieve a comprehensive moral understanding. This research underscores the limitations of current PLMs in processing cross-cultural differences in values and highlights the importance of developing culturally aware AI systems that better align with universal human values.

This study investigates the internal reasoning mechanism of language models during symbolic multi-step reasoning, motivated by the question of whether chain-of-thought (CoT) outputs are faithful to the model's internals. Specifically, we inspect when they internally determine their answers, particularly before or after CoT begins, to determine whether models follow a post-hoc "think-to-talk" mode or a step-by-step "talk-to-think" mode of explanation. Through causal probing experiments in controlled arithmetic reasoning tasks, we found systematic internal reasoning patterns across models; for example, simple subproblems are solved before CoT begins, and more complicated multi-hop calculations are performed during CoT.

Large language models (LLMs) have become increasingly pivotal in various domains due the recent advancements in their performance capabilities. However, concerns persist regarding biases in LLMs, including gender, racial, and cultural biases derived from their training data. These biases raise critical questions about the ethical deployment and societal impact of LLMs. Acknowledging these concerns, this study investigates whether LLMs accurately reflect cross-cultural variations and similarities in moral perspectives. In assessing whether the chosen LLMs capture patterns of divergence and agreement on moral topics across cultures, three main methods are employed: (1) comparison of model-generated and survey-based moral score variances, (2) cluster alignment analysis to evaluate the correspondence between country clusters derived from model-generated moral scores and those derived from survey data, and (3) probing LLMs with direct comparative prompts. All three methods involve the use of systematic prompts and token pairs designed to assess how well LLMs understand and reflect cultural variations in moral attitudes. The findings of this study indicate overall variable and low performance in reflecting cross-cultural differences and similarities in moral values across the models tested, highlighting the necessity for improving models' accuracy in capturing these nuances effectively. The insights gained from this study aim to inform discussions on the ethical development and deployment of LLMs in global contexts, emphasizing the importance of mitigating biases and promoting fair representation across diverse cultural perspectives.

Deep neural network (DNN)-based policy models like vision-language-action (VLA) models are transformative in automating complex decision-making across applications by interpreting multi-modal data. However, scaling these models greatly increases computational costs, which presents challenges in fields like robot manipulation and autonomous driving that require quick, accurate responses. To address the need for deployment on resource-limited hardware, we propose a new quantization framework for IL-based policy models that fine-tunes parameters to enhance robustness against low-bit precision errors during training, thereby maintaining efficiency and reliability under constrained conditions. Our evaluations with representative robot manipulation for 4-bit weight-quantization on a real edge GPU demonstrate that our framework achieves up to 2.5x speedup and 2.5x energy savings while preserving accuracy. For 4-bit weight and activation quantized self-driving models, the framework achieves up to 3.7x speedup and 3.1x energy saving on a low-end GPU. These results highlight the practical potential of deploying IL-based policy models on resource-constrained devices.

This project addresses challenges posed by the widespread use of abbreviations and acronyms in digital texts. We propose a novel method that combines document preprocessing, regular expressions, and a large language model to identify abbreviations and map them to their corresponding expansions. The regular expressions alone are often insufficient to extract expansions, at which point our approach leverages GPT-4 to analyze the text surrounding the acronyms. By limiting the analysis to only a small portion of the surrounding text, we mitigate the risk of obtaining incorrect or multiple expansions for an acronym. There are several known challenges in processing text with acronyms, including polysemous acronyms, non-local and ambiguous acronyms. Our approach enhances the precision and efficiency of NLP techniques by addressing these issues with automated acronym identification and disambiguation. This study highlights the challenges of working with PDF files and the importance of document preprocessing. Furthermore, the results of this work show that neither regular expressions nor GPT-4 alone can perform well. Regular expressions are suitable for identifying acronyms but have limitations in finding their expansions within the paper due to a variety of formats used for expressing acronym-expansion pairs and the tendency of authors to omit expansions within the text. GPT-4, on the other hand, is an excellent tool for obtaining expansions but struggles with correctly identifying all relevant acronyms. Additionally, GPT-4 poses challenges due to its probabilistic nature, which may lead to slightly different results for the same input. Our algorithm employs preprocessing to eliminate irrelevant information from the text, regular expressions for identifying acronyms, and a large language model to help find acronym expansions to provide the most accurate and consistent results.

LLMs demand significant computational resources for both pre-training and fine-tuning, requiring distributed computing capabilities due to their large model sizes \cite{sastry2024computing}. Their complex architecture poses challenges throughout the entire AI lifecycle, from data collection to deployment and monitoring \cite{OECD_AIlifecycle}. Addressing critical AI system challenges, such as explainability, corrigibility, interpretability, and hallucination, necessitates a systematic methodology and rigorous benchmarking \cite{guldimann2024complai}. To effectively improve AI systems, we must precisely identify systemic vulnerabilities through quantitative evaluation, bolstering system trustworthiness. The enactment of the EU AI Act \cite{EUAIAct} by the European Parliament on March 13, 2024, establishing the first comprehensive EU-wide requirements for the development, deployment, and use of AI systems, further underscores the importance of tools and methodologies such as Z-Inspection. It highlights the need to enrich this methodology with practical benchmarks to effectively address the technical challenges posed by AI systems. To this end, we have launched a project that is part of the AI Safety Bulgaria initiatives \cite{AI_Safety_Bulgaria}, aimed at collecting and categorizing AI benchmarks. This will enable practitioners to identify and utilize these benchmarks throughout the AI system lifecycle.

Errors in understanding visual information in images (i.e., visual perception errors) remain a major source of mistakes in Large Vision Language Models (LVLMs). While further analysis is essential, there is a deficiency in datasets for evaluating the visual perception of LVLMs. In this work, we introduce VisOnlyQA, a new dataset designed to directly evaluate the visual perception capabilities of LVLMs on questions about geometric and numerical information in scientific figures. Our dataset enables us to analyze the visual perception of LVLMs for fine-grained visual information, independent of other capabilities such as reasoning. The evaluation set of VisOnlyQA includes 1,200 multiple-choice questions in 12 tasks on four categories of figures. We also provide synthetic training data consisting of 70k instances. Our experiments on VisOnlyQA highlight the following findings: (i) 20 LVLMs we evaluate, including GPT-4o and Gemini 1.5 Pro, work poorly on the visual perception tasks in VisOnlyQA, while human performance is nearly perfect. (ii) Fine-tuning on synthetic training data demonstrates the potential for enhancing the visual perception of LVLMs, but observed improvements are limited to certain tasks and specific models. (iii) Stronger language models improve the visual perception of LVLMs. In summary, our experiments suggest that both training data and model architectures should be improved to enhance the visual perception capabilities of LVLMs. The datasets, code, and model responses are provided at https://github.com/psunlpgroup/VisOnlyQA.

Evaluations of Large Language Models (LLMs) on knowledge-intensive tasks and factual accuracy often focus on high-resource languages primarily because datasets for low-resource languages (LRLs) are scarce. In this paper, we present Uhura -- a new benchmark that focuses on two tasks in six typologically-diverse African languages, created via human translation of existing English benchmarks. The first dataset, Uhura-ARC-Easy, is composed of multiple-choice science questions. The second, Uhura-TruthfulQA, is a safety benchmark testing the truthfulness of models on topics including health, law, finance, and politics. We highlight the challenges creating benchmarks with highly technical content for LRLs and outline mitigation strategies. Our evaluation reveals a significant performance gap between proprietary models such as GPT-4o and o1-preview, and Claude models, and open-source models like Meta's LLaMA and Google's Gemma. Additionally, all models perform better in English than in African languages. These results indicate that LMs struggle with answering scientific questions and are more prone to generating false claims in low-resource African languages. Our findings underscore the necessity for continuous improvement of multilingual LM capabilities in LRL settings to ensure safe and reliable use in real-world contexts. We open-source the Uhura Benchmark and Uhura Platform to foster further research and development in NLP for LRLs.

Language Agent could be endowed with different mechanisms for autonomous task accomplishment. Current agents typically rely on fixed mechanisms or a set of mechanisms activated in a predefined order, limiting their adaptation to varied potential task solution structures. To this end, this paper proposes \textbf{A}daptive \textbf{L}anguage \textbf{A}gent \textbf{M}echanism \textbf{A}ctivation Learning with Self-Exploration (\textbf{ALAMA}), which focuses on optimizing mechanism activation adaptability without reliance on expert models. Initially, it builds a harmonized agent framework (\textbf{UniAct}) to \textbf{Uni}fy different mechanisms via \textbf{Act}ions. Then it leverages a training-efficient optimization method based on self-exploration to enable the UniAct to adaptively activate the appropriate mechanisms according to the potential characteristics of the task. Experimental results demonstrate significant improvements in downstream agent tasks, affirming the effectiveness of our approach in facilitating more dynamic and context-sensitive mechanism activation.

Large Language Models (LLMs) demonstrate remarkable capabilities in various reasoning tasks. However, they encounter significant challenges when it comes to scientific reasoning, particularly in physics, which requires not only mathematical reasoning but also factual and conceptual understanding. When addressing complex physics problems, LLMs typically face three key issues: problem miscomprehension, incorrect concept application, and computational errors. While each of these problems can be addressed individually, there is a need for a generalized approach that can tackle all three issues simultaneously. To address this, we introduce Mixture of Refinement Agents (MoRA), a novel agentic refinement framework that iteratively refines the LLM generated base solution by correcting the aforementioned errors, resulting in a significant performance improvement for open-source LLMs. Our approach aims to bridge the gap between opensource LLMs and GPT-4o by utilizing the latter as error identifier to guide these refinement agents. We evaluate our approach on the SciEval and MMLU subsets along with our own physics dataset (PhysicsQA). MoRA significantly improves the performance of Llama-3-70B and Gemma-2-27B on these datasets, achieving up to a 16% increase in final answer accuracy.