An official website of the United States government
Code LLMs have the potential to make it easier for non-experts to understand and write code. However, current CodeLLM benchmarks rely on a single expert-written prompt per problem, making it hard to generalize their success to non-expert users. In this paper, we present a new natural-language-to-code benchmark of prompts written by a key population of non-experts: beginning programmers. StudentEval contains 1,749 prompts written by 80 students who have only completed one introductory Python course. StudentEval contains numerous non-expert prompts describing the same problem, enabling exploration of key factors in prompt success. We use StudentEval to evaluate 12 Code LLMs and find that StudentEval is a better discriminator of model performance than existing benchmarks. Our analysis of student prompting strategies reveals that nondeterministic LLM sampling can mislead students about the quality of their descriptions, a finding with key implications for Code LLMs in education.
more »
« less
This content will become publicly available on April 29, 2026
Substance Beats Style: Why Beginning Students Fail to Code with LLMs
Although LLMs are increasing the productivity of professional programmers, existing work shows that beginners struggle to prompt LLMs to solve text-to-code tasks (Nguyen et al., 2024; Prather et al., 2024; Mordechai et al., 2024). Why is this the case? This paper explores two competing hypotheses about the cause of student-LLM miscommunication: (1) students simply lack the technical vocabulary needed to write good prompts, and (2) students do not understand the extent of information that LLMs need to solve code generation tasks. We study (1) with a causal intervention experiment on technical vocabulary and (2) by analyzing graphs that abstract how students edit prompts and the different failures that they encounter. We find that substance beats style: a poor grasp of technical vocabulary is merely correlated with prompt failure; that the information content of prompts predicts success; that students get stuck making trivial edits; and more. Our findings have implications for the use of LLMs in programming education, and for efforts to make computing more accessible with LLMs.
more »
« less
- Award ID(s):
- 2326175
- PAR ID:
- 10599017
- Publisher / Repository:
- Association for Computational Linguistics
- Date Published:
- Volume:
- 1
- Page Range / eLocation ID:
- 8541 to 8610
- Format(s):
- Medium: X
- Location:
- Albuquerque, New Mexico
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Large language models (LLM) are perceived to offer promising potentials for automating security tasks, such as those found in security operation centers (SOCs). As a first step towards evaluating this perceived potential, we investigate the use of LLMs in software pentesting, where the main task is to automatically identify software security vulnerabilities in source code. We hypothesize that an LLM-based AI agent can be improved over time for a specific security task as human operators interact with it. Such improvement can be made, as a first step, by engineering prompts fed to the LLM based on the responses produced, to include relevant contexts and structures so that the model provides more accurate results. Such engineering efforts become sustainable if the prompts that are engineered to produce better results on current tasks, also produce better results on future unknown tasks. To examine this hypothesis, we utilize the OWASP Benchmark Project 1.2 which contains 2,740 hand-crafted source code test cases containing various types of vulnerabilities. We divide the test cases into training and testing data, where we engineer the prompts based on the training data (only), and evaluate the final system on the testing data. We compare the AI agent’s performance on the testing data against the performance of the agent without the prompt engineering. We also compare the AI agent’s results against those from SonarQube, a widely used static code analyzer for security testing. We built and tested multiple versions of the AI agent using different off-the-shelf LLMs – Google’s Gemini-pro, as well as OpenAI’s GPT-3.5-Turbo and GPT-4-Turbo (with both chat completion and assistant APIs). The results show that using LLMs is a viable approach to build an AI agent for software pentesting that can improve through repeated use and prompt engineering.more » « less
-
This study used three pairs of problem-posing tasks to examine the impact of different prompts on students’ problem posing. Two kinds of prompts were involved. The first asked students to pose 2–3 different mathematical problems without specifying other requirements for the problems, whereas the second kind of prompt did specify additional requirements. A total of 2124 students’ responses were analyzed to examine the impact of the prompts along multiple dimensions. In response to problem-posing prompts with more specific requirements, students tended to engage in more in-depth mathematical thinking and posed much more linguistically and semantically complex problems with more relationships or steps required to solve them. The findings from this study not only contribute to our understanding of problem-posing processes but also have direct implications for teaching mathematics through problem posing.more » « less
-
Large language models offer new ways of empowering people to program robot applications-namely, code generation via prompting. However, the code generated by LLMs is susceptible to errors. This work reports a preliminary exploration that empirically characterizes common errors produced by LLMs in robot programming. We categorize these errors into two phases: interpretation and execution. In this work, we focus on errors in execution and observe that they are caused by LLMs being “forgetful” of key information provided in user prompts. Based on this observation, we propose prompt engineering tactics designed to reduce errors in execution. We then demonstrate the effectiveness of these tactics with three language models: ChatGPT, Bard, and LLaMA-2. Finally, we discuss lessons learned from using LLMs in robot programming and call for the benchmarking of LLM-powered end-user development of robot applications.more » « less
-
Language models produce a distribution over the next token; can we use this information to recover the prompt tokens? We consider the problem of language model inversion and show that next-token probabilities contain a surprising amount of information about the preceding text. Often we can recover the text in cases where it is hidden from the user, motivating a method for recovering unknown prompts given only the model's current distribution output. We consider a variety of model access scenarios, and show how even without predictions for every token in the vocabulary we can recover the probability vector through search. On Llama-2 7b, our inversion method reconstructs prompts with a BLEU of 59 and token-level F1 of 78 and recovers 27% of prompts exactly. Code for reproducing all experiments is available at this http URL.more » « less
