Education systems are dynamically changing to accommodate technological advances, industrial and societal needs, and to enhance students' learning journeys. Curriculum specialists and educators constantly revise taught subjects across educational grades to identify gaps, introduce new learning topics, and enhance the learning outcomes. This process is usually done within the same subjects (e.g. math) or across related subjects (e.g. math and physics) considering the same and different educational levels, leading to massive multi-layer comparisons. Having nuanced data about subjects, topics, and learning outcomes structured within a dataset, empowers us to leverage data science to better understand the progression of various learning topics. In this paper, Bidirectional Encoder Representations from Transformers (BERT) topic modeling was used to extract topics from the curriculum, which were then used to identify relationships between subjects, track their progression, and identify conceptual gaps. We found that grouping learning outcomes by common topics helped specialists reduce redundancy and introduce new concepts in the curriculum. We built a dashboard to avail the methodology to curriculum specials. Finally, we tested the validity of the approach with subject matter experts.
We investigate efficient methods for training Large Language Models (LLMs) to possess capabilities in multiple specialized domains, such as coding, math reasoning and world knowledge. Our method, named Branch-Train-MiX (BTX), starts from a seed model, which is branched to train experts in embarrassingly parallel fashion with high throughput and reduced communication cost. After individual experts are asynchronously trained, BTX brings together their feedforward parameters as experts in Mixture-of-Expert (MoE) layers and averages the remaining parameters, followed by an MoE-finetuning stage to learn token-level routing. BTX generalizes two special cases, the Branch-Train-Merge method, which does not have the MoE finetuning stage to learn routing, and sparse upcycling, which omits the stage of training experts asynchronously. Compared to alternative approaches, BTX achieves the best accuracy-efficiency tradeoff.
To thoroughly assess the mathematical reasoning abilities of Large Language Models (LLMs), we need to carefully curate evaluation datasets covering diverse mathematical concepts and mathematical problems at different difficulty levels. In pursuit of this objective, we propose FineMath in this paper, a fine-grained mathematical evaluation benchmark dataset for assessing Chinese LLMs. FineMath is created to cover the major key mathematical concepts taught in elementary school math, which are further divided into 17 categories of math word problems, enabling in-depth analysis of mathematical reasoning abilities of LLMs. All the 17 categories of math word problems are manually annotated with their difficulty levels according to the number of reasoning steps required to solve these problems. We conduct extensive experiments on a wide range of LLMs on FineMath and find that there is still considerable room for improvements in terms of mathematical reasoning capability of Chinese LLMs. We also carry out an in-depth analysis on the evaluation process and methods that have been overlooked previously. These two factors significantly influence the model results and our understanding of their mathematical reasoning capabilities. The dataset will be publicly available soon.
Despite the effectiveness of data selection for large language models (LLMs) during pretraining and instruction fine-tuning phases, improving data efficiency in supervised fine-tuning (SFT) for specialized domains poses significant challenges due to the complexity of fine-tuning data. To bridge this gap, we introduce an effective and scalable data selection method for SFT, SmallToLarge (S2L), which leverages training trajectories from small models to guide the data selection for larger models. We demonstrate through extensive experiments that S2L significantly improves data efficiency in SFT for mathematical problem-solving, reducing the training data to just 11% of the original MathInstruct dataset (Yue et al., 2023) to match full dataset performance while outperforming state-of-the-art data selection algorithms by an average of 4.7% across 6 in- and out-domain evaluation datasets. Remarkably, selecting only 50K data for SFT, S2L achieves a 32.7% accuracy on the most challenging MATH (Hendrycks et al., 2021) benchmark, improving Phi-2 (Li et al., 2023b) by 16.6%. In clinical text summarization on the MIMIC-III dataset (Johnson et al., 2016), S2L again outperforms training on the full dataset using only 50% of the data. Notably, S2L can perform data selection using a reference model 40x smaller than the target model, proportionally reducing the cost of data selection.
Mathematical capabilities were previously believed to emerge in common language models only at a very large scale or require extensive math-related pre-training. This paper shows that the LLaMA-2 7B model with common pre-training already exhibits strong mathematical abilities, as evidenced by its impressive accuracy of 97.7% and 72.0% on the GSM8K and MATH benchmarks, respectively, when selecting the best response from 256 random generations. The primary issue with the current base model is the difficulty in consistently eliciting its inherent mathematical capabilities. Notably, the accuracy for the first answer drops to 49.5% and 7.9% on the GSM8K and MATH benchmarks, respectively. We find that simply scaling up the SFT data can significantly enhance the reliability of generating correct answers. However, the potential for extensive scaling is constrained by the scarcity of publicly available math questions. To overcome this limitation, we employ synthetic data, which proves to be nearly as effective as real data and shows no clear saturation when scaled up to approximately one million samples. This straightforward approach achieves an accuracy of 82.6% on GSM8K and 40.6% on MATH using LLaMA-2 7B models, surpassing previous models by 14.2% and 20.8%, respectively. We also provide insights into scaling behaviors across different reasoning complexities and error types.
Generative AI technologies such as ChatGPT, Gemini, and MidJourney have made remarkable progress in recent years. Recent literature has documented ChatGPT's positive impact on productivity in areas where it has strong expertise, attributable to extensive training datasets, such as the English language and Python/SQL programming. However, there is still limited literature regarding ChatGPT's performance in areas where its capabilities could still be further enhanced. This paper aims to fill this gap. We conducted an experiment in which economics students were asked to perform writing analysis tasks in a non-English language (specifically, Thai) and math & data analysis tasks using a less frequently used programming package (specifically, Stata). The findings suggest that, on average, participants performed better using ChatGPT in terms of scores and time taken to complete the tasks. However, a detailed examination reveals that 34% of participants saw no improvement in writing analysis tasks, and 42% did not improve in math & data analysis tasks when employing ChatGPT. Further investigation indicated that higher-ability students, as proxied by their econometrics grades, were the ones who performed worse in writing analysis tasks when using ChatGPT. We also found evidence that students with better digital skills performed better with ChatGPT. This research provides insights on the impact of generative AI. Thus, stakeholders can make informed decisions to implement appropriate policy frameworks or redesign educational systems. It also highlights the critical role of human skills in addressing and complementing the limitations of technology.
Large language models (LLMs) are highly effective in various natural language processing (NLP) tasks. However, they are susceptible to producing unreliable conjectures in ambiguous contexts called hallucination. This paper presents a new method for evaluating LLM hallucination in Question Answering (QA) based on the unanswerable math word problem (MWP). To support this approach, we innovatively develop a dataset called Unanswerable Math Word Problem (UMWP) which comprises 5200 questions across five categories. We developed an evaluation methodology combining text similarity and mathematical expression detection to determine whether LLM considers the question unanswerable. The results of extensive experiments conducted on 31 LLMs, including GPT-3, InstructGPT, LLaMA, and Claude, demonstrate that in-context learning and reinforcement learning with human feedback (RLHF) training significantly enhance the model's ability to avoid hallucination. We show that utilizing MWP is a reliable and effective approach to assess hallucination. Our code and data are available at https://github.com/Yuki-Asuuna/UMWP.
The African continent lacks enough qualified teachers which hampers the provision of adequate learning support. An AI could potentially augment the efforts of the limited number of teachers, leading to better learning outcomes. Towards that end, this work describes and evaluates the first key output for the NSMQ AI Grand Challenge, which proposes a robust, real-world benchmark for such an AI: "Build an AI to compete live in Ghana's National Science and Maths Quiz (NSMQ) competition and win - performing better than the best contestants in all rounds and stages of the competition". The NSMQ is an annual live science and mathematics competition for senior secondary school students in Ghana in which 3 teams of 2 students compete by answering questions across biology, chemistry, physics, and math in 5 rounds over 5 progressive stages until a winning team is crowned for that year. In this work, we built Brilla AI, an AI contestant that we deployed to unofficially compete remotely and live in the Riddles round of the 2023 NSMQ Grand Finale, the first of its kind in the 30-year history of the competition. Brilla AI is currently available as a web app that livestreams the Riddles round of the contest, and runs 4 machine learning systems: (1) speech to text (2) question extraction (3) question answering and (4) text to speech that work together in real-time to quickly and accurately provide an answer, and then say it with a Ghanaian accent. In its debut, our AI answered one of the 4 riddles ahead of the 3 human contesting teams, unofficially placing second (tied). Improvements and extensions of this AI could potentially be deployed to offer science tutoring to students and eventually enable millions across Africa to have one-on-one learning interactions, democratizing science education.
Large language models (LLMs) have demonstrated remarkable capabilities in problem-solving. However, their proficiency in solving mathematical problems remains inadequate. We propose MathScale, a simple and scalable method to create high-quality mathematical reasoning data using frontier LLMs (e.g., {\tt GPT-3.5}). Inspired by the cognitive mechanism in human mathematical learning, it first extracts topics and knowledge points from seed math questions and then build a concept graph, which is subsequently used to generate new math questions. MathScale exhibits effective scalability along the size axis of the math dataset that we generate. As a result, we create a mathematical reasoning dataset (MathScaleQA) containing two million math question-answer pairs. To evaluate mathematical reasoning abilities of LLMs comprehensively, we construct {\sc MwpBench}, a benchmark of Math Word Problems, which is a collection of ten datasets (including GSM8K and MATH) covering K-12, college, and competition level math problems. We apply MathScaleQA to fine-tune open-source LLMs (e.g., LLaMA-2 and Mistral), resulting in significantly improved capabilities in mathematical reasoning. Evaluated on {\sc MwpBench}, MathScale-7B achieves state-of-the-art performance across all datasets, surpassing its best peers of equivalent size by 42.9\% in micro average accuracy and 43.7\% in macro average accuracy, respectively.
Creating effective educational materials generally requires expensive and time-consuming studies of student learning outcomes. To overcome this barrier, one idea is to build computational models of student learning and use them to optimize instructional materials. However, it is difficult to model the cognitive processes of learning dynamics. We propose an alternative approach that uses Language Models (LMs) as educational experts to assess the impact of various instructions on learning outcomes. Specifically, we use GPT-3.5 to evaluate the overall effect of instructional materials on different student groups and find that it can replicate well-established educational findings such as the Expertise Reversal Effect and the Variability Effect. This demonstrates the potential of LMs as reliable evaluators of educational content. Building on this insight, we introduce an instruction optimization approach in which one LM generates instructional materials using the judgments of another LM as a reward function. We apply this approach to create math word problem worksheets aimed at maximizing student learning gains. Human teachers' evaluations of these LM-generated worksheets show a significant alignment between the LM judgments and human teacher preferences. We conclude by discussing potential divergences between human and LM opinions and the resulting pitfalls of automating instructional design.
Large language models (LLMs) have shown great potential in complex reasoning tasks, yet their performance is often hampered by the scarcity of high-quality and reasoning-focused training datasets. Addressing this challenge, we propose Key-Point-Driven Data Synthesis (KPDDS), a novel data synthesis framework that synthesizes question-answer pairs by leveraging key points and exemplar practices from authentic data sources. KPDDS ensures the generation of novel questions with rigorous quality control and substantial scalability. As a result, we present KPMath, an extensive synthetic dataset tailored for mathematical reasoning, comprising over 800K question-answer pairs. Utilizing KPMath and augmenting it with additional reasoning-intensive corpora, we create the comprehensive KPMath-Plus dataset. The Qwen1.5-72B model, fine-tuned on KPMath-Plus, achieves 87.0% PASS@1 accuracy on GSM8K and 58.3% on MATH, surpassing competitors in the 7B to 70B range and best commercial models like GPT-4 across multiple math reasoning datasets.
Highlighting particularly relevant regions of an image can improve the performance of vision-language models (VLMs) on various vision-language (VL) tasks by guiding the model to attend more closely to these regions of interest. For example, VLMs can be given a "visual prompt", where visual markers such as bounding boxes delineate key image regions. However, current VLMs that can incorporate visual guidance are either proprietary and expensive or require costly training on curated data that includes visual prompts. We introduce Contrastive Region Guidance (CRG), a training-free guidance method that enables open-source VLMs to respond to visual prompts. CRG contrasts model outputs produced with and without visual prompts, factoring out biases revealed by the model when answering without the information required to produce a correct answer (i.e., the model's prior). CRG achieves substantial improvements in a wide variety of VL tasks: When region annotations are provided, CRG increases absolute accuracy by up to 11.1% on ViP-Bench, a collection of six diverse region-based tasks such as recognition, math, and object relationship reasoning. We also show CRG's applicability to spatial reasoning, with 10% improvement on What'sUp, as well as to compositional generalization -- improving accuracy by 11.5% and 7.5% on two challenging splits from SugarCrepe -- and to image-text alignment for generated images, where we improve by up to 8.4 AUROC and 6.8 F1 points on SeeTRUE. When reference regions are absent, CRG allows us to re-rank proposed regions in referring expression comprehension and phrase grounding benchmarks like RefCOCO/+/g and Flickr30K Entities, with an average gain of 3.2% in accuracy. Our analysis explores alternative masking strategies for CRG, quantifies CRG's probability shift, and evaluates the role of region guidance strength, empirically validating CRG's design choices.
In reasoning tasks, even a minor error can cascade into inaccurate results, leading to suboptimal performance of large language models in such domains. Earlier fine-tuning approaches sought to mitigate this by leveraging more precise supervisory signals from human labeling, larger models, or self-sampling, although at a high cost. Conversely, we develop a method that avoids external resources, relying instead on introducing perturbations to the input. Our training approach randomly masks certain tokens within the chain of thought, a technique we found to be particularly effective for reasoning tasks. When applied to fine-tuning with GSM8K on Llama-2-7B, this method achieved a 5\% improvement in GSM8K accuracy and a 10\% improvement in GSM-IC accuracy over standard supervised fine-tuning with a few codes modified. Furthermore, it is complementary to existing methods. When integrated with related explicit data augmentation methods, it leads to improvements across five datasets of various augmentation methods, as well as two different base models. We further investigate the mechanisms behind this improvement through case studies and quantitative analysis, suggesting that our approach may provide superior support for the model in capturing long-distance dependencies, especially those related to questions. This enhancement could deepen understanding of the premises in questions and prior steps. Our code is available at Github.
Automatically generating feedback via large language models (LLMs) in intelligent tutoring systems and online learning platforms has the potential to improve the learning outcomes of many students. However, both feedback generation and evaluation are challenging: feedback content has to be valid especially in subjects like math, which requires models to understand the problem, the solution, and where the student's error lies. Feedback also has to be pedagogically valid to reflect effective tutoring strategies, such as explaining possible misconceptions and encouraging the student, among other desirable features. In this work, we address both problems of automatically generating and evaluating feedback while considering both correctness and alignment. First, we propose a rubric for evaluating math feedback and show that GPT-4 is able to effectively use it to annotate human-written and LLM-generated feedback. Second, we propose a framework for feedback generation that optimizes both correctness and alignment using reinforcement learning (RL). Specifically, we use GPT-4's annotations to create preferences over feedback pairs in an augmented dataset for training via direct preference optimization (DPO). We show that our methods significantly increase the correctness and alignment of generated feedback with Llama 2, an open-source LLM, qualitatively analyze our generation and evaluation systems using case studies, and outline several areas for future work.
Although large language models demonstrate emergent abilities in solving math word problems, there is a challenging task in complex multi-step mathematical reasoning tasks. To improve model performance on mathematical reasoning tasks, previous work has conducted supervised fine-tuning on open-source models by improving the quality and quantity of data. In this paper, we propose a novel approach, named Brain, to imitate human thought processes to enhance mathematical reasoning abilities, using the Frontal Lobe Model to generate plans, and then employing the Parietal Lobe Model to generate code and execute to obtain answers. First, we achieve SOTA performance in comparison with Code LLaMA 7B based models through this method. Secondly, we find that plans can be explicitly extracted from natural language, code, or formal language. Our code and data are publicly available at https://github.com/cyzhh/Brain.
Large language models (LLMs) have achieved impressive performance across various mathematical reasoning benchmarks. However, there are increasing debates regarding whether these models truly understand and apply mathematical knowledge or merely rely on shortcuts for mathematical reasoning. One essential and frequently occurring evidence is that when the math questions are slightly changed, LLMs can behave incorrectly. This motivates us to evaluate the robustness of LLMs' math reasoning capability by testing a wide range of question variations. We introduce the adversarial grade school math (GSM-Plus) dataset, an extension of GSM8K augmented with various mathematical perturbations. Our experiments on 25 LLMs and 4 prompting techniques show that while LLMs exhibit different levels of math reasoning abilities, their performances are far from robust. In particular, even for problems that have been solved in GSM8K, LLMs can make mistakes when new statements are added or the question targets are altered. We also explore whether more robust performance can be achieved by composing existing prompting methods, in which we try an iterative method that generates and verifies each intermediate thought based on its reasoning goal and calculation result.
Recently, large language models (LLMs) have garnered widespread attention for their exceptional capabilities. Prompts are central to the functionality and performance of LLMs, making them highly valuable assets. The increasing reliance on high-quality prompts has driven significant growth in prompt services. However, this growth also expands the potential for prompt leakage, increasing the risk that attackers could replicate original functionalities, create competing products, and severely infringe on developers' intellectual property. Despite these risks, prompt leakage in real-world prompt services remains underexplored. In this paper, we present PRSA, a practical attack framework designed for prompt stealing. PRSA infers the detailed intent of prompts through very limited input-output analysis and can successfully generate stolen prompts that replicate the original functionality. Extensive evaluations demonstrate PRSA's effectiveness across two main types of real-world prompt services. Specifically, compared to previous works, it improves the attack success rate from 17.8% to 46.1% in prompt marketplaces and from 39% to 52% in LLM application stores, respectively. Notably, in the attack on "Math", one of the most popular educational applications in OpenAI's GPT Store with over 1 million conversations, PRSA uncovered a hidden Easter egg that had not been revealed previously. Besides, our analysis reveals that higher mutual information between a prompt and its output correlates with an increased risk of leakage. This insight guides the design and evaluation of two potential defenses against the security threats posed by PRSA. We have reported these findings to the prompt service vendors, including PromptBase and OpenAI, and actively collaborate with them to implement defensive measures.
Large Language Model (LLM)-based agents have shown effectiveness across many applications. However, their use in data science scenarios requiring solving long-term interconnected tasks, dynamic data adjustments and domain expertise remains challenging. Previous approaches primarily focus on individual tasks, making it difficult to assess the complete data science workflow. Moreover, they struggle to handle real-time changes in intermediate data and fail to adapt dynamically to evolving task dependencies inherent to data science problems. In this paper, we present Data Interpreter, an LLM-based agent designed to automatically solve various data science problems end-to-end. Our Data Interpreter incorporates two key modules: 1) Hierarchical Graph Modeling, which breaks down complex problems into manageable subproblems, enabling dynamic node generation and graph optimization; and 2) Programmable Node Generation, a technique that refines and verifies each subproblem to iteratively improve code generation results and robustness. Extensive experiments consistently demonstrate the superiority of Data Interpreter. On InfiAgent-DABench, it achieves a 25% performance boost, raising accuracy from 75.9% to 94.9%. For machine learning and open-ended tasks, it improves performance from 88% to 95%, and from 60% to 97%, respectively. Moreover, on the MATH dataset, Data Interpreter achieves remarkable performance with a 26% improvement compared to state-of-the-art baselines. The code is available at https://github.com/geekan/MetaGPT.
We propose a framework for robust evaluation of reasoning capabilities of language models, using functional variants of benchmarks. Models that solve a reasoning test should exhibit no difference in performance over the static version of a problem compared to a snapshot of the functional variant. We have rewritten the relevant fragment of the MATH benchmark into its functional variant MATH(), with functionalization of other benchmarks to follow. When evaluating current state-of-the-art models over snapshots of MATH(), we find a reasoning gap -- the percentage difference between the static and functional accuracies. We find reasoning gaps from 58.35% to 80.31% among the state-of-the-art closed and open weights models that perform well on static benchmarks, with the caveat that the gaps are likely to be smaller with more sophisticated prompting strategies. Here we show that models which anecdotally have good reasoning performance over real-world tasks, have quantifiable lower gaps, motivating the open problem of building "gap 0" models. Code for evaluation and new evaluation datasets, three MATH() snapshots, are publicly available at https://github.com/consequentai/fneval/.
The BigCode project, an open-scientific collaboration focused on the responsible development of Large Language Models for Code (Code LLMs), introduces StarCoder2. In partnership with Software Heritage (SWH), we build The Stack v2 on top of the digital commons of their source code archive. Alongside the SWH repositories spanning 619 programming languages, we carefully select other high-quality data sources, such as GitHub pull requests, Kaggle notebooks, and code documentation. This results in a training set that is 4x larger than the first StarCoder dataset. We train StarCoder2 models with 3B, 7B, and 15B parameters on 3.3 to 4.3 trillion tokens and thoroughly evaluate them on a comprehensive set of Code LLM benchmarks. We find that our small model, StarCoder2-3B, outperforms other Code LLMs of similar size on most benchmarks, and also outperforms StarCoderBase-15B. Our large model, StarCoder2- 15B, significantly outperforms other models of comparable size. In addition, it matches or outperforms CodeLlama-34B, a model more than twice its size. Although DeepSeekCoder- 33B is the best-performing model at code completion for high-resource languages, we find that StarCoder2-15B outperforms it on math and code reasoning benchmarks, as well as several low-resource languages. We make the model weights available under an OpenRAIL license and ensure full transparency regarding the training data by releasing the SoftWare Heritage persistent IDentifiers (SWHIDs) of the source code data.