Due to the cost-prohibitive nature of training Large Language Models (LLMs), fine-tuning has emerged as an attractive alternative for specializing LLMs for specific tasks using limited compute resources in a cost-effective manner. In this paper, we characterize sparse Mixture of Experts (MoE) based LLM fine-tuning to understand their accuracy and runtime performance on a single GPU. Our evaluation provides unique insights into the training efficacy of sparse and dense versions of MoE models, as well as their runtime characteristics, including maximum batch size, execution time breakdown, end-to-end throughput, GPU hardware utilization, and load distribution. Our study identifies the optimization of the MoE layer as crucial for further improving the performance of LLM fine-tuning. Using our profiling results, we also develop and validate an analytical model to estimate the cost of LLM fine-tuning on the cloud. This model, based on parameters of the model and GPU architecture, estimates LLM throughput and the cost of training, aiding practitioners in industry and academia to budget the cost of fine-tuning a specific model.

The vast portion of workloads employing LLMs involves answering questions grounded on PDF or scan content. We introduce the Arctic-TILT achieving accuracy on par with models 1000$\times$ its size on these use cases. It can be fine-tuned and deployed on a single 24GB GPU, lowering operational costs while processing Visually Rich Documents with up to 400k tokens. The model establishes state-of-the-art results on seven diverse Document Understanding benchmarks, as well as provides reliable confidence scores and quick inference, which are essential for processing files in large-scale or time-sensitive enterprise environments.

This paper explores the question of how accurately current large language models can perform logical reasoning in natural language, with an emphasis on whether these models exhibit reasoning biases similar to humans. Specifically, our study focuses on syllogistic reasoning, a form of deductive reasoning extensively studied in cognitive science as a natural form of human reasoning. We present a syllogism dataset called NeuBAROCO, which consists of syllogistic reasoning problems in English and Japanese. This dataset was originally designed for psychological experiments to assess human reasoning capabilities using various forms of syllogisms. Our experiments with leading large language models indicate that these models exhibit reasoning biases similar to humans, along with other error tendencies. Notably, there is significant room for improvement in reasoning problems where the relationship between premises and hypotheses is neither entailment nor contradiction. We also present experimental results and in-depth analysis using a new Chain-of-Thought prompting method, which asks LLMs to translate syllogisms into abstract logical expressions and then explain their reasoning process. Our analysis using this method suggests that the primary limitations of LLMs lie in the reasoning process itself rather than the interpretation of syllogisms.

The importance of recommender systems is growing rapidly due to the exponential increase in the volume of content generated daily. This surge in content presents unique challenges for designing effective recommender systems. Key among these challenges is the need to effectively leverage the vast amounts of natural language data and images that represent user preferences. This paper presents a novel approach to enhancing recommender systems by leveraging Large Language Models (LLMs) and deep learning techniques. The proposed framework aims to improve the accuracy and relevance of recommendations by incorporating multi-modal information processing and by the use of unified latent space representation. The study explores the potential of LLMs to better understand and utilize natural language data in recommendation contexts, addressing the limitations of previous methods. The framework efficiently extracts and integrates text and image information through LLMs, unifying diverse modalities in a latent space to simplify the learning process for the ranking model. Experimental results demonstrate the enhanced discriminative power of the model when utilizing multi-modal information. This research contributes to the evolving field of recommender systems by showcasing the potential of LLMs and multi-modal data integration to create more personalized and contextually relevant recommendations.

The Artificial Intelligence Act of the European Union mandates that providers and deployers of high-risk AI systems establish a quality management system (QMS). Among other criteria, a QMS shall help to i) identify, analyze, evaluate, and mitigate risks, ii) ensure evidence of compliance with training, validation, and testing data, and iii) verify and document the AI system design and quality. Current research mainly addresses conceptual considerations and framework designs for AI risk assessment and auditing processes. However, it often overlooks practical tools that actively involve and support humans in checking and documenting high-risk or general-purpose AI systems. This paper addresses this gap by proposing requirements derived from legal regulations and a generic design and architecture of a QMS for AI systems verification and documentation. A first version of a prototype QMS is implemented, integrating LLMs as examples of AI systems and focusing on an integrated risk management sub-service. The prototype is evaluated on i) a user story-based qualitative requirements assessment using potential stakeholder scenarios and ii) a technical assessment of the required GPU storage and performance.

Developing questions that are pedagogically sound, relevant, and promote learning is a challenging and time-consuming task for educators. Modern-day large language models (LLMs) generate high-quality content across multiple domains, potentially helping educators to develop high-quality questions. Automated educational question generation (AEQG) is important in scaling online education catering to a diverse student population. Past attempts at AEQG have shown limited abilities to generate questions at higher cognitive levels. In this study, we examine the ability of five state-of-the-art LLMs of different sizes to generate diverse and high-quality questions of different cognitive levels, as defined by Bloom's taxonomy. We use advanced prompting techniques with varying complexity for AEQG. We conducted expert and LLM-based evaluations to assess the linguistic and pedagogical relevance and quality of the questions. Our findings suggest that LLms can generate relevant and high-quality educational questions of different cognitive levels when prompted with adequate information, although there is a significant variance in the performance of the five LLms considered. We also show that automated evaluation is not on par with human evaluation.

The Eye Movements on Machine-Generated Texts Corpus (EMTeC) is a naturalistic eye-movements-while-reading corpus of 107 native English speakers reading machine-generated texts. The texts are generated by three large language models using five different decoding strategies, and they fall into six different text type categories. EMTeC entails the eye movement data at all stages of pre-processing, i.e., the raw coordinate data sampled at 2000 Hz, the fixation sequences, and the reading measures. It further provides both the original and a corrected version of the fixation sequences, accounting for vertical calibration drift. Moreover, the corpus includes the language models' internals that underlie the generation of the stimulus texts: the transition scores, the attention scores, and the hidden states. The stimuli are annotated for a range of linguistic features both at text and at word level. We anticipate EMTeC to be utilized for a variety of use cases such as, but not restricted to, the investigation of reading behavior on machine-generated text and the impact of different decoding strategies; reading behavior on different text types; the development of new pre-processing, data filtering, and drift correction algorithms; the cognitive interpretability and enhancement of language models; and the assessment of the predictive power of surprisal and entropy for human reading times. The data at all stages of pre-processing, the model internals, and the code to reproduce the stimulus generation, data pre-processing and analyses can be accessed via https://github.com/DiLi-Lab/EMTeC/.

As large language models continue to scale up, the imperative for fault tolerance in distributed deep learning systems intensifies, becoming a focal area of AI infrastructure research. Checkpoint has emerged as the predominant fault tolerance strategy, with extensive studies dedicated to optimizing its efficiency. However, the advent of the sparse Mixture-of-Experts (MoE) model presents new challenges for traditional checkpoint techniques due to the substantial increase in model size, despite comparable computational demands to dense models. Breaking new ground in the realm of efficient fault tolerance for MoE model training, we introduce a novel Partial Experts Checkpoint (PEC) mechanism alongside a corresponding PEC fault-tolerant system. Our approach strategically checkpoints a selected subset of experts, thereby significantly reducing the checkpoint size for MoE models to a level comparable with that of dense models. The empirical analysis on our 8-expert GPT-MoE model demonstrates that the proposed PEC approach facilitates a substantial 54.2% decrease in the size of non-redundant checkpoint (no data-parallel duplication), without compromising the final model quality. Moreover, our PEC fault-tolerant system achieves a 76.9% reduction in checkpoint workload per data-parallel distributed rank, thereby correspondingly diminishing the checkpointing time and facilitating complete overlap with the training process.

Large language models (LLMs) have exhibited remarkable proficiency across a diverse array of natural language processing (NLP) tasks. However, adapting LLMs to downstream applications typically necessitates computationally intensive and memory-demanding fine-tuning procedures. To mitigate these burdens, parameter-efficient fine-tuning (PEFT) techniques have emerged as a promising approach to tailor LLMs with minimal computational overhead. While PEFT methods offer substantial advantages, they do not fully address the pervasive issue of bias propagation from pre-training data. In this work, we introduce Bias-Aware Low-Rank Adaptation (BA-LoRA), a novel PEFT method designed to counteract bias inheritance. BA-LoRA incorporates three distinct regularization terms: (1) consistency regularizer, (2) diversity regularizer, and (3) singular vector decomposition regularizer. These regularizers collectively aim to improve the generative models' consistency, diversity, and generalization capabilities during the fine-tuning process. Through extensive experiments on a variety of natural language understanding (NLU) and natural language generation (NLG) tasks, employing prominent LLMs such as LLaMA, Mistral, and Gemma, we demonstrate that BA-LoRA surpasses the performance of LoRA and its state-of-the-art variants. Moreover, our method effectively mitigates the deleterious effects of pre-training bias, leading to more reliable and robust model outputs. The code is available at https://github.com/cyp-jlu-ai/BA-LoRA.

Large language models (LLMs) are supposed to acquire unconscious human knowledge and feelings, such as social common sense and biases, by training models from large amounts of text. However, it is not clear how much the sentiments of specific social groups can be captured in various LLMs. In this study, we focus on social groups defined in terms of nationality, religion, and race/ethnicity, and validate the extent to which sentiments between social groups can be captured in and extracted from LLMs. Specifically, we input questions regarding sentiments from one group to another into LLMs, apply sentiment analysis to the responses, and compare the results with social surveys. The validation results using five representative LLMs showed higher correlations with relatively small p-values for nationalities and religions, whose number of data points were relatively large. This result indicates that the LLM responses including the inter-group sentiments align well with actual social survey results.