Yuheng Hu

Sorouralsadat Fatemi, Yuheng Hu

While Large Language Models (LLMs) have shown impressive capabilities in numerous Natural Language Processing (NLP) tasks, they still struggle with financial question answering (QA), particularly when numerical reasoning is required. Recently, LLM-based multi-agent frameworks have demonstrated remarkable effectiveness in multi-step reasoning, which is crucial for financial QA tasks as it involves extracting relevant information from tables and text and then performing numerical reasoning on the extracted data to infer answers. In this study, we propose a multi-agent framework incorporating a critic agent that reflects on the reasoning steps and final answers for each question. Additionally, we enhance our system by adding multiple critic agents, each focusing on a specific aspect of the answer. Our results indicate that this framework significantly improves performance compared to single-agent reasoning, with an average performance increase of 15% for the LLaMA3-8B model and 5% for the LLaMA3-70B model. Furthermore, our framework performs on par with, and in some cases surpasses, larger single-agent LLMs such as LLaMA3.1-405B and GPT-4o-mini, though it falls slightly short compared to Claude-3.5 Sonnet. Overall, our framework presents an effective solution to enhance open-source LLMs for financial QA tasks, offering a cost-effective alternative to larger models like Claude-3.5 Sonnet.

Yaoyu Chen, Yuheng Hu, Yingda Lu

Prevailing large language models (LLMs) are capable of human responses simulation through its unprecedented content generation and reasoning abilities. However, it is not clear whether and how to leverage LLMs to simulate field experiments. In this paper, we propose and evaluate two prompting strategies: the observer mode that allows a direct prediction on main conclusions and the participant mode that simulates distributions of responses from participants. Using this approach, we examine fifteen well cited field experimental papers published in INFORMS and MISQ, finding encouraging alignments between simulated experimental results and the actual results in certain scenarios. We further identify topics of which LLMs underperform, including gender difference and social norms related research. Additionally, the automatic and standardized workflow proposed in this paper enables the possibility of a large-scale screening of more papers with field experiments. This paper pioneers the utilization of large language models (LLMs) for simulating field experiments, presenting a significant extension to previous work which focused solely on lab environments. By introducing two novel prompting strategies, observer and participant modes, we demonstrate the ability of LLMs to both predict outcomes and replicate participant responses within complex field settings. Our findings indicate a promising alignment with actual experimental results in certain scenarios, achieving a stimulation accuracy of 66% in observer mode. This study expands the scope of potential applications for LLMs and illustrates their utility in assisting researchers prior to engaging in expensive field experiments. Moreover, it sheds light on the boundaries of LLMs when used in simulating field experiments, serving as a cautionary note for researchers considering the integration of LLMs into their experimental toolkit.

Sorouralsadat Fatemi, Yuheng Hu

Financial trading has been a challenging task, as it requires the integration of vast amounts of data from various modalities. Traditional deep learning and reinforcement learning methods require large training data and often involve encoding various data types into numerical formats for model input, which limits the explainability of model behavior. Recently, LLM-based agents have demonstrated remarkable advancements in handling multi-modal data, enabling them to execute complex, multi-step decision-making tasks while providing insights into their thought processes. This research introduces a multi-modal multi-agent system designed specifically for financial trading tasks. Our framework employs a team of specialized LLM-based agents, each adept at processing and interpreting various forms of financial data, such as textual news reports, candlestick charts, and trading signal charts. A key feature of our approach is the integration of a reflection module, which conducts analyses of historical trading signals and their outcomes. This reflective process is instrumental in enhancing the decision-making capabilities of the system for future trading scenarios. Furthermore, the ablation studies indicate that the visual reflection module plays a crucial role in enhancing the decision-making capabilities of our framework.

Sorouralsadat Fatemi, Yuheng Hu

Financial sentiment analysis plays a crucial role in uncovering latent patterns and detecting emerging trends, enabling individuals to make well-informed decisions that may yield substantial advantages within the constantly changing realm of finance. Recently, Large Language Models (LLMs) have demonstrated their effectiveness in diverse domains, showcasing remarkable capabilities even in zero-shot and few-shot in-context learning for various Natural Language Processing (NLP) tasks. Nevertheless, their potential and applicability in the context of financial sentiment analysis have not been thoroughly explored yet. To bridge this gap, we employ two approaches: in-context learning (with a focus on gpt-3.5-turbo model) and fine-tuning LLMs on a finance-domain dataset. Given the computational costs associated with fine-tuning LLMs with large parameter sizes, our focus lies on smaller LLMs, spanning from 250M to 3B parameters for fine-tuning. We then compare the performances with state-of-the-art results to evaluate their effectiveness in the finance-domain. Our results demonstrate that fine-tuned smaller LLMs can achieve comparable performance to state-of-the-art fine-tuned LLMs, even with models having fewer parameters and a smaller training dataset. Additionally, the zero-shot and one-shot performance of LLMs produces comparable results with fine-tuned smaller LLMs and state-of-the-art outcomes. Furthermore, our analysis demonstrates that there is no observed enhancement in performance for finance-domain sentiment analysis when the number of shots for in-context learning is increased.

Sorouralsadat Fatemi, Yuheng Hu, Maryam Mousavi

Large Language Models (LLMs) have demonstrated impressive capabilities across diverse Natural Language Processing (NLP) tasks, including language understanding, reasoning, and generation. However, general-domain LLMs often struggle with financial tasks due to the technical and specialized nature of financial texts. This study investigates the efficacy of instruction fine-tuning smaller-scale LLMs, including Mistral-7B, Llama3-8B, and Phi3-mini, to enhance their performance in financial text classification tasks. We fine-tuned both instruction-tuned and base models across four financial classification tasks, achieving significant improvements in task-specific performance. Furthermore, we evaluated the zero-shot capabilities of these fine-tuned models on three unseen complex financial tasks, including argument classification, deal completeness classification, and causal classification. Our results indicate while base model fine-tuning led to greater degradation, instruction-tuned models maintained more robust performance. To address this degradation, we employed model merging techniques, integrating single-task domain-specific fine-tuned models with the base model. Using this merging method resulted in significant enhancements in zero-shot performance, even exceeding the original model's accuracy on certain datasets. Our findings underscore the effectiveness of instruction fine-tuning and model merging for adapting LLMs to specialized financial text classification tasks.

Yuheng Hu, Shelly D. Farnham, Andres Monroy-Hernandez

Social media systems promise powerful opportunities for people to connect to timely, relevant information at the hyper local level. Yet, finding the meaningful signal in noisy social media streams can be quite daunting to users. In this paper, we present and evaluate Whoo.ly, a web service that provides neighborhood-specific information based on Twitter posts that were automatically inferred to be hyperlocal. Whoo.ly automatically extracts and summarizes hyperlocal information about events, topics, people, and places from these Twitter posts. We provide an overview of our design goals with Whoo.ly and describe the system including the user interface and our unique event detection and summarization algorithms. We tested the usefulness of the system as a tool for finding neighborhood information through a comprehensive user study. The outcome demonstrated that most participants found Whoo.ly easier to use than Twitter and they would prefer it as a tool for exploring their neighborhoods.

Yuheng Hu, Ajita John, Fei Wang et al.

Social media channels such as Twitter have emerged as popular platforms for crowds to respond to public events such as speeches, sports and debates. While this promises tremendous opportunities to understand and make sense of the reception of an event from the social media, the promises come entwined with significant technical challenges. In particular, given an event and an associated large scale collection of tweets, we need approaches to effectively align tweets and the parts of the event they refer to. This in turn raises questions about how to segment the event into smaller yet meaningful parts, and how to figure out whether a tweet is a general one about the entire event or specific one aimed at a particular segment of the event. In this work, we present ET-LDA, an effective method for aligning an event and its tweets through joint statistical modeling of topical influences from the events and their associated tweets. The model enables the automatic segmentation of the events and the characterization of tweets into two categories: (1) episodic tweets that respond specifically to the content in the segments of the events, and (2) steady tweets that respond generally about the events. We present an efficient inference method for this model, and a comprehensive evaluation of its effectiveness over existing methods. In particular, through a user study, we demonstrate that users find the topics, the segments, the alignment, and the episodic tweets discovered by ET-LDA to be of higher quality and more interesting as compared to the state-of-the-art, with improvements in the range of 18-41%.

Yuheng Hu, Sushovan De, Yi Chen et al.

Data Cleaning is a long standing problem, which is growing in importance with the mass of uncurated web data. State of the art approaches for handling inconsistent data are systems that learn and use conditional functional dependencies (CFDs) to rectify data. These methods learn data patterns--CFDs--from a clean sample of the data and use them to rectify the dirty/inconsistent data. While getting a clean training sample is feasible in enterprise data scenarios, it is infeasible in web databases where there is no separate curated data. CFD based methods are unfortunately particularly sensitive to noise; we will empirically demonstrate that the number of CFDs learned falls quite drastically with even a small amount of noise. In order to overcome this limitation, we propose a fully probabilistic framework for cleaning data. Our approach involves learning both the generative and error (corruption) models of the data and using them to clean the data. For generative models, we learn Bayes networks from the data. For error models, we consider a maximum entropy framework for combing multiple error processes. The generative and error models are learned directly from the noisy data. We present the details of the framework and demonstrate its effectiveness in rectifying web data.