Human Programmer

Research on "human programmer" currently centers on understanding and enhancing the collaboration between human developers and large language models (LLMs) like GPT-4 in software development. Key areas of investigation include assessing the quality and bias of LLM-generated code and comments, analyzing differences in coding styles between humans and LLMs, and developing methods to improve the efficiency and effectiveness of human-AI pair programming. This research aims to optimize software development workflows, potentially increasing programmer productivity and improving code quality through the strategic integration of AI tools.

Papers

June 8, 2023

June 3, 2023

Towards Coding Social Science Datasets with Language Models
Christopher Michael Rytting, Taylor Sorensen, Lisa Argyle, Ethan Busby, Nancy Fulda, Joshua Gubler, David Wingate
Language Model Social Science Human Programmer Synthetic Instruction Human Accuracy

June 2, 2023

Do Large Language Models Pay Similar Attention Like Human Programmers When Generating Code?
Bonan Kou, Shengmai Chen, Zhijie Wang, Lei Ma, Tianyi Zhang
Code Generation Attention Pattern Code Generation Benchmark Human Programmer Code Generator User Attention

June 1, 2023

Analysis of ChatGPT on Source Code
Ahmed R. Sadik, Antonello Ceravola, Frank Joublin, Jibesh Patra
General Analysis ChatGPT Generated Conversation Code Generation Source Code Human Programmer Code Analysis Programming Community

May 27, 2023

A Two-Stage Decoder for Efficient ICD Coding
Thanh-Tung Nguyen, Viktor Schlegel, Abhinav Kashyap, Stefan Winkler
Classification Code Multi Label Text Classification Human Programmer International Classification of Disease Medical Object Detection

May 24, 2023

The Larger They Are, the Harder They Fail: Language Models do not Recognize Identifier Swaps in Python
Antonio Valerio Miceli-Barone, Fazl Barez, Ioannis Konstas, Shay B. Cohen
Language Model Multiplicative Size Scaling Code Generation Task Human Programmer Identity Collision

May 10, 2023

Humans are Still Better than ChatGPT: Case of the IEEEXtreme Competition
Anis Koubaa, Basit Qureshi, Adel Ammar, Zahid Khan, Wadii Boulila, Lahouari Ghouti
ChatGPT Generated Conversation Real Human Case Relevance Human Programmer Programming Language

April 13, 2023

LeafAI: query generator for clinical cohort discovery rivaling a human programmer
Nicholas J Dobbins, Bin Han, Weipeng Zhou, Kristine Lan, H. Nina Kim, Robert Harrington, Ozlem Uzuner, Meliha Yetisgen
Model Agnostic Clinical Data Query Generation Human Programmer Unified Medical Language System Clinical Trial Eligibility Criterion

October 30, 2022

Medical Codes Prediction from Clinical Notes: From Human Coders to Machines
Byung-Hak Kim
New Machine Clinical Note Unstructured Clinical Code Model Human Programmer Medical Coding Medical Code Prediction

October 29, 2022

Aligning Offline Metrics and Human Judgments of Value for Code Generation Models
Victor Dibia, Adam Fourney, Gagan Bansal, Forough Poursabzi-Sangdeh, Han Liu, Saleema Amershi
Net Present Value Human Judgment Code Generation Model Human Programmer Offline Evaluation Syntactic Relation Working Programmer AI Pair Functional Correctness

October 25, 2022

Leveraging Open Data and Task Augmentation to Automated Behavioral Coding of Psychotherapy Conversations in Low-Resource Scenarios
Zhuohao Chen, Nikolaos Flemotomos, Zac E. Imel, David C. Atkins, Shrikanth Narayanan
Cross Domain Behavior Model Open Data Human Programmer Low Resource Scenario Psychotherapy Session Analogy Task

June 30, 2022

GitHub Copilot AI pair programmer: Asset or Liability?
Arghavan Moradi Dakhel, Vahid Majdinasab, Amin Nikanjam, Foutse Khomh, Michel C. Desmarais, Zhen Ming, Jiang
Program Synthesis Programming Task Human Programmer NextGen Communication COPILOT Microsoft Copilot Liability Inducing Text Individual Production Asset AI Pair

June 24, 2022

A Test for Evaluating Performance in Human-Computer Systems
Andres Campero, Michelle Vaccaro, Jaeyoon Song, Haoran Wen, Abdullah Almaatouq, Thomas W. Malone
System Performance Level Test Performance Improvement Human Machine Interaction Human Programmer

May 13, 2022

Productivity Assessment of Neural Code Completion
Albert Ziegler, Eirini Kalliamvakou, Shawn Simister, Ganesh Sittampalam, Alice Li, Andrew Rice, Devon Rifkin, Edward Aftandilian
Real World Code Performance Evaluation Software Development Human Programmer Snippet Extraction Neural Code Completion

Human Programmer

Papers

Is AI the better programming partner? Human-Human Pair Programming vs. Human-AI pAIr Programming

When to Show a Suggestion? Integrating Human Feedback in AI-Assisted Programming

Towards Coding Social Science Datasets with Language Models

Do Large Language Models Pay Similar Attention Like Human Programmers When Generating Code?

Analysis of ChatGPT on Source Code

A Two-Stage Decoder for Efficient ICD Coding

The Larger They Are, the Harder They Fail: Language Models do not Recognize Identifier Swaps in Python

Humans are Still Better than ChatGPT: Case of the IEEEXtreme Competition

LeafAI: query generator for clinical cohort discovery rivaling a human programmer

Medical Codes Prediction from Clinical Notes: From Human Coders to Machines

Aligning Offline Metrics and Human Judgments of Value for Code Generation Models

Leveraging Open Data and Task Augmentation to Automated Behavioral Coding of Psychotherapy Conversations in Low-Resource Scenarios

GitHub Copilot AI pair programmer: Asset or Liability?

A Test for Evaluating Performance in Human-Computer Systems

Productivity Assessment of Neural Code Completion