Spreadsheets are a vital tool for end-user data management. Using large language models for formula authoring assistance in these environments can be difficult, as these models are expensive to train and challenging to deploy due to their size (up to billions of parameters). We present FLAME, a transformer-based model trained exclusively on Excel formulas that leverages domain insights to achieve competitive performance while being substantially smaller (60M parameters) and training on two orders of magnitude less data. We curate a training dataset using sketch deduplication, introduce an Excel-specific formula tokenizer, and use domain-specific versions of masked span prediction and noisy auto-encoding as pre-training objectives. We evaluate FLAME on formula repair, formula completion, and similarity-based formula retrieval. FLAME can outperform much larger models, such as the Davinci (175B) and Cushman (12B) variants of Codex and CodeT5 (220M), in 10 of 14 evaluation settings for the repair and completion tasks. For formula retrieval, FLAME outperforms CodeT5, CodeBERT, and GraphCodeBERT.

Most programmers make mistakes when writing code. Some of these mistakes are small and require few edits to the original program – a class of errors recently termed last mile mistakes. These errors break the flow for experienced developers and can stump novice programmers. Existing automated repair techniques targeting this class of errors are language-specific and do not easily carry over to new languages. Transferring symbolic approaches requires substantial engineering and neural approaches require data and retraining. We introduce RING, a multilingual repair engine powered by a large language model trained on code (LLMC) such as Codex. Such a multilingual engine enables a flipped model for programming assistance, one where the programmer writes code and the AI assistance suggests fixes, compared to traditional code suggestion technology. Taking inspiration from the way programmers manually fix bugs, we show that a prompt-based strategy that conceptualizes repair as localization, transformation, and candidate ranking, can successfully repair programs in multiple languages with minimal effort. We present the first results for such a multilingual repair engine by evaluating on 6 different languages and comparing performance to language-specific repair engines. We show that RING can outperform language-specific repair engines for three of these languages.

Most users of low-code platforms, such as Excel and PowerApps, write programs in domain-specific formula languages to carry out nontrivial tasks. Often users can write most of the program they want, but introduce small mistakes that yield broken formulas. These mistakes, which can be both syntactic and semantic, are hard for low-code users to identify and fix, even though they can be resolved with just a few edits. We formalize the problem of producing such edits as the last-mile repair problem. To address this problem, we developed LaMirage, a LAst-MIle RepAir-engine GEnerator that combines symbolic and neural techniques to perform last-mile repair in low-code formula languages. LaMirage takes a grammar and a set of domain-specific constraints/rules, which jointly approximate the target language, and uses these to generate a repair engine that can fix formulas in that language. To tackle the challenges of localizing errors and ranking candidate repairs, LaMirage leverages neural techniques, whereas it relies on symbolic methods to generate candidate edits. This combination allows LaMirage to find repairs that satisfy the provided grammar and constraints, and then pick the most natural repair. We compare LaMirage to state-of-the-art neural and symbolic approaches on 400 real Excel and Power Fx formulas, where LaMirage outperforms all baselines. We release these benchmarks to encourage subsequent work in low-code domains.

Recent development in NL2CODE (Natural Language to Code) research allows end-users, especially novice programmers to create a concrete implementation of their ideas such as data visualization by providing natural language (NL) instructions. An NL2CODE system often fails to achieve its goal due to three major challenges: the user's words have contextual semantics, the user may not include all details needed for code generation, and the system results are imperfect and require further refinement. To address the aforementioned three challenges for NL to Visualization, we propose a new approach and its supporting tool named NL2VIZ with three salient features: (1) leveraging not only the user's NL input but also the data and program context that the NL query is upon, (2) using hard/soft constraints to reflect different confidence levels in the constraints retrieved from the user input and data/program context, and (3) providing support for result refinement and reuse. We implement NL2VIZ in the Jupyter Notebook environment and evaluate NL2VIZ on a real-world visualization benchmark and a public dataset to show the effectiveness of NL2VIZ. We also conduct a user study involving 6 data scientist professionals to demonstrate the usability of NL2VIZ, the readability of the generated code, and NL2VIZ's effectiveness in helping users generate desired visualizations effectively and efficiently.

We formalize and study “programming by rewards” (PBR), a new approach for specifying and synthesizing subroutines for optimizing some quantitative metric such as performance, resource utilization, or correctness over a benchmark. A PBR specification consists of (1) input features x, and (2) a reward function r, modeled as a black-box component (which we can only run), that assigns a reward for each execution. The goal of the synthesizer is to synthesize a "decision function" f which transforms the features to a decision value for the black-box component so as to maximize the expected reward E[r ∘ f (x)] for executing decisions f(x) for various values of x. We consider a space of decision functions in a DSL of loop-free if-then-else programs, which can branch on linear functions of the input features in a tree-structure and compute a linear function of the inputs in the leaves of the tree. We find that this DSL captures decision functions that are manually written in practice by programmers. Our technical contribution is the use of continuous-optimization techniques to perform synthesis of such decision functions as if-then-else programs. We also show that the framework is theoretically-founded —in cases when the rewards satisfy nice properties, the synthesized code is optimal in a precise sense. We have leveraged PBR to synthesize non-trivial decision functions related to search and ranking heuristics in the PROSE codebase (an industrial strength program synthesis framework) and achieve competitive results to manually written procedures over multiple man years of tuning. We present empirical evaluation against other baseline techniques over real-world case studies (including PROSE) as well on simple synthetic benchmarks.

Data repositories often consist of text files in a wide variety of standard formats, ad-hoc formats, as well as mixtures of formats where data in one format is embedded into a different format. It is therefore a significant challenge to parse these files into a structured tabular form, which is important to enable any downstream data processing. We present Unravel, an extensible framework for structure interpretation of ad-hoc formats. Unravel can automatically, with no user input, extract tabular data from a diverse range of standard, ad-hoc and mixed format files. The framework is also easily extensible to add support for previously unseen formats, and also supports interactivity from the user in terms of examples to guide the system when specialized data extraction is desired. Our key insight is to allow arbitrary combination of extraction and parsing techniques through a concept called partial structures. Partial structures act as a common language through which the file structure can be shared and refined by different techniques. This makes Unravel more powerful than applying the individual techniques in parallel or sequentially. Further, with this rule-based extensible approach, we introduce the novel notion of re-interpretation where the variety of techniques supported by our system can be exploited to improve accuracy while optimizing for particular quality measures or restricted environments. On our benchmark of 617 text files gathered from a variety of sources, Unravel is able to extract the intended table in many more cases compared to state-of-the-art techniques.

Providing feedback on programming assignments is a tedious task for the instructor, and even impossible in large Massive Open Online Courses with thousands of students. Previous research has suggested that program repair techniques can be used to generate feedback in programming education. In this paper, we present a novel fully automated program repair algorithm for introductory programming assignments. The key idea of the technique, which enables automation and scalability, is to use the existing correct student solutions to repair the incorrect attempts. We evaluate the approach in two experiments: (I) We evaluate the number, size and quality of the generated repairs on 4,293 incorrect student attempts from an existing MOOC. We find that our approach can repair 97% of student attempts, while 81% of those are small repairs of good quality. (II) We conduct a preliminary user study on performance and repair usefulness in an interactive teaching setting. We obtain promising initial results (the average usefulness grade 3.4 on a scale from 1 to 5), and conclude that our approach can be used in an interactive setting.

Formal frameworks for cost analysis of programs have been widely studied in the unary setting and, to a limited extent, in the relational setting. However, many of these frameworks focus only on the cost aspect, largely side-lining functional properties that are often a prerequisite for cost analysis, thus leaving many interesting programs out of their purview. In this paper, we show that elegant, simple, expressive proof systems combining cost analysis and functional properties can be built by combining already known ingredients: higher-order refinements and cost monads. Specifically, we derive two syntax-directed proof systems, UC and RC, for unary and relational cost analysis, by adding a cost monad to a (syntax-directed) logic of higher-order programs. We study the metatheory of the systems, show that several nontrivial examples can be verified in them, and prove that existing frameworks for cost analysis (RelCost and RAML) can be embedded in them.

Providing feedback on programming assignments manually is a tedious, error prone, and time-consuming task. In this paper, we motivate and address the problem of generating feedback on performance aspects in introductory programming assignments. We studied a large number of functionally correct student solutions to introductory programming assignments and observed: (1) There are different algorithmic strategies, with varying levels of efficiency, for solving a given problem. These different strategies merit different feedback. (2) The same algorithmic strategy can be implemented in countless different ways, which are not relevant for reporting feedback on the student program. We propose a light-weight programming language extension that allows a teacher to define an algorithmic strategy by specifying certain key values that should occur during the execution of an implementation. We describe a dynamic analysis based approach to test whether a student's program matches a teacher's specification. Our experimental results illustrate the effectiveness of both our specification language and our dynamic analysis. On one of our benchmarks consisting of 2316 functionally correct implementations to 3 programming problems, we identified 16 strategies that we were able to describe using our specification language (in 95 minutes after inspecting 66, i.e., around 3%, implementations). Our dynamic analysis correctly matched each implementation with its corresponding specification, thereby automatically producing the intended feedback.