Research Projects on Large Language Model and Natural Language Processing

From Blind Solvers to Logical Thinkers: Benchmarking LLMs' Logical Integrity on Faulty Mathematical Problems

Consider the math problem: "Lily received 3 cookies from her best friend yesterday and ate 5 for breakfast. Today, her friend gave her 3 more cookies. How many cookies does Lily have now?" Many large language models (LLMs) in previous research approach this problem by calculating the answer "1" using the equation "3 - 5 + 3." However, from a human perspective, we recognize the inherent flaw in this problem: Lily cannot eat 5 cookies if she initially only had 3. This discrepancy prompts a key question: Are current LLMs merely Blind Solver that apply mathematical operations without deeper reasoning, or can they function as Logical Thinker capable of identifying logical inconsistencies? To explore this question, we propose a benchmark dataset, FaultyMath, which includes faulty math problems of rich diversity: i) multiple mathematical categories, e.g., algebra, geometry, number theory, etc., ii) varying levels of difficulty, and iii) different origins of faultiness -- ranging from violations of common sense and ambiguous statements to mathematical contradictions and more. We evaluate a broad spectrum of LLMs, including open-source, closed-source, and math-specialized models, using FaultyMath across three dimensions: (i) How accurately can the models detect faulty math problems without being explicitly prompted to do so? (ii) When provided with hints -- either correct or misleading -- about the validity of the problems, to what extent do LLMs adapt to become reliable Logical Thinker? (iii) How trustworthy are the explanations generated by LLMs when they recognize a math problem as flawed? Through extensive experimentation and detailed analysis, our results demonstrate that existing LLMs largely function as Blind Solver and fall short of the reasoning capabilities required to perform as Logical Thinker.

FaultyMath

Assessing the Creativity of LLMs in Proposing Novel Solutions to Mathematical Problems

The mathematical capabilities of AI systems are complex and multifaceted. Most existing research has predominantly focused on the correctness of AI-generated solutions to mathematical problems. In this work, we argue that beyond producing correct answers, AI systems should also be capable of, or assist humans in, developing novel solutions to mathematical challenges. This study explores the creative potential of Large Language Models (LLMs) in mathematical reasoning, an aspect that has received limited attention in prior research. We introduce a novel framework and benchmark, CreativeMath, which encompasses problems ranging from middle school curricula to Olympic-level competitions, designed to assess LLMs' ability to propose innovative solutions after some known solutions have been provided. Our experiments demonstrate that, while LLMs perform well on standard mathematical tasks, their capacity for creative problem-solving varies considerably. Notably, the Gemini-1.5-Pro model outperformed other LLMs in generating novel solutions. This research opens a new frontier in evaluating AI creativity, shedding light on both the strengths and limitations of LLMs in fostering mathematical innovation, and setting the stage for future developments in AI-assisted mathematical discovery.

CreativeMath

Data Augmentation for Text Classification with EASE

In image classification, efficient data augmentation (DA) is easy with cropping, rotating, blurring etc. It works because a cropped/blurred “cat” is still a “cat”. In other words, the augmented sample does not require additional labeling as in most cases the augmented sample retains the original label. In text classification (TC) existing methods have random insertion, deletion of random words or punctuation, but the semantics change very easily. Therefore, these methods that use the same label for the augmented samples simply inject more noise in many cases. Moreover, acquiring new labels for the augmented sample requires training on the original data first to get a good estimate of the new labels. In this work, we present EASE, a simple but dependable DA technique for TC that has four easy steps: Extract Units, Acquire Labels, Sift and Employ. We extract meaningful units as augmented samples from original data samples and use powerful tools to acquire labels for them before they are sifted and merged. Previous DA techniques, like EDA and AEDA, excel with sequential models but struggle with transformer-based models that heavily rely on token order. EASE, in contrast, performs well with these models, demonstrating stability, speed, and minimal adverse effects. We tested our intuitive method on multiple challenging datasets sensitive to augmentation, and experimental results have indicated the efficacy of DA with EASE.

ease

LLM on DataFrame Question Answering Without Data Exposure

We introduce DataFrame question answering (QA), a novel task that utilizes natural language processing (NLP) models to generate Pandas queries for information retrieval and data analysis on dataframes, emphasizing safe and non-revealing data handling. Specifically, our method, leveraging large language model (LLM), which solely relies on dataframe column names, not only ensures data privacy but also significantly reduces the context window in the prompt, streamlining information processing and addressing major challenges in LLM-based data analysis. We propose DataFrame QA as a comprehensive framework that includes safe Pandas query generation and code execution. Various LLMs are evaluated on the renowned WikiSQL dataset and our newly developed UCI-DataFrameQA, tailored for complex data analysis queries. Our findings indicate that GPT-4 performs well on both datasets, underscoring its capability in securely retrieving and aggregating dataframe values and conducting sophisticated data analyses. This approach, deployable in a zero-shot manner without prior training or adjustments, proves to be highly adaptable and secure for diverse applications.

dataframeqa