Association for Artificial Intelligence 2025, Alahi, Alexandre, Martorella, Tommaso et al. · Underline Science Inc. · 2025
Conditioning image generation facilitates seamless editing and the creation of photorealistic images. However, conditioning on noisy or Out-of-Distribution (OoD) images poses significant challenges, particularly in balancing fidelity to the input and realism of the output. We introduce Confident Ordinary Differential Editing (CODE), a novel approach for image synthesis that effectively handles OoD guidance images. Utilizing a diffusion model as a generative prior, CODE enhances images through score-based updates along the probability-flow Ordinary Differential Equation (ODE) trajectory. This method requires no task-specific training, handcrafted modules, or assumptions, and is compatible with any diffusion model. Positioned at the intersection of conditional image generation and blind image restoration, CODE operates in a fully blind manner, relying solely on a pre-trained generative model. Our method introduces an alternative approach to blind restoration: instead of targeting a specific ground truth image based on assumptions about the underlying corruption, CODE aims to increase the likelihood of the input image while maintaining fidelity. This results in the most probable in-distribution image around the input. Our contributions are twofold. First, CODE introduces a novel editing method based on ODE providing enhanced control, realism, and fidelity compared to SDE-based counterpart. Second, we introduce a confidence interval-based clipping method, which improves CODE’s effectiveness by allowing it to disregard certain pixels or information, thus enhancing the restoration process in a blind manner. Experimental results demonstrate CODE’s effectiveness over existing methods, particularly in scenarios involving severe degradation or OoD inputs.
Cited by 0Open access
Éric Deneault · IntechOpen eBooks · 2025
Abstract From the discovery of DNA to the development of highly precise and versatile molecular tools to modify it, gene editing has evolved to radically transform how we fight cancer and genetic diseases. This chapter describes key milestones that led to CRISPR-based technologies becoming specific tools for genome editing. Techniques based on Cas9 nuclease, base editing, prime editing, and click editing now enable accurate ex vivo and in vivo gene corrections. The expansion to other CRISPR orthologs allows for more diverse genetic modifications. These breakthroughs offer hope for addressing both monogenic and complex disorders. Despite constant progress, ongoing research is required to manage off-target effects and collateral damage that may co-occur during the gene-editing process.
Cited by 0Open access
Elliot Bendoly · Cambridge University Press eBooks · 2008
Many effective decision support systems rely not only on the ability of a manager to present information, analysis, and meaningful dynamics (e.g., through graphics), but also on enabling users to realize the intended use of those elements by themselves (without the developer holding their hand).
Cited by 0
Youngseok Yoon, Christian Kästner (cmu Isr · 2015
clusions or recommendations expressed in this material are those of the author and do not necessarily reflect those of KFAS or the National Science Foundation. ii
Cited by 0
Bastien van Delft, Tommaso Martorella, Alexandre Alahi · arXiv (Cornell University) · 2024
Conditioning image generation facilitates seamless editing and the creation of photorealistic images. However, conditioning on noisy or Out-of-Distribution (OoD) images poses significant challenges, particularly in balancing fidelity to the input and realism of the output. We introduce Confident Ordinary Differential Editing (CODE), a novel approach for image synthesis that effectively handles OoD guidance images. Utilizing a diffusion model as a generative prior, CODE enhances images through score-based updates along the probability-flow Ordinary Differential Equation (ODE) trajectory. This method requires no task-specific training, no handcrafted modules, and no assumptions regarding the corruptions affecting the conditioning image. Our method is compatible with any diffusion model. Positioned at the intersection of conditional image generation and blind image restoration, CODE operates in a fully blind manner, relying solely on a pre-trained generative model. Our method introduces an alternative approach to blind restoration: instead of targeting a specific ground truth image based on assumptions about the underlying corruption, CODE aims to increase the likelihood of the input image while maintaining fidelity. This results in the most probable in-distribution image around the input. Our contributions are twofold. First, CODE introduces a novel editing method based on ODE, providing enhanced control, realism, and fidelity compared to its SDE-based counterpart. Second, we introduce a confidence interval-based clipping method, which improves CODE's effectiveness by allowing it to disregard certain pixels or information, thus enhancing the restoration process in a blind manner. Experimental results demonstrate CODE's effectiveness over existing methods, particularly in scenarios involving severe degradation or OoD inputs.
Cited by 0Open access
Li, Weichen, Jan, Albert, Ray, Baishakhi et al. · ArXiv.org · 2025
Code editing is a foundational task in software development, where its effectiveness depends on whether it introduces desired code property changes without changing the original code's intended functionality. Existing approaches often formulate code editing as an implicit end-to-end task, omitting the fact that code-editing procedures inherently consist of discrete and explicit steps. Thus, they suffer from suboptimal performance and lack of robustness and generalization. We introduce EditLord, a code editing framework that makes the code transformation steps explicit. Our key insight is to employ a language model (LM) as an inductive learner to extract code editing rules from the training code pairs as concise meta-rule sets. Such rule sets will be manifested for each training sample to augment them for finetuning or assist in prompting- and iterative-based code editing. EditLord outperforms the state-of-the-art by an average of 22.7% in editing performance and 58.1% in robustness while achieving 20.2% higher functional correctness across critical software engineering and security applications, LM models, and editing modes.
Cited by 0Open access
Abhay patle · Zenodo (CERN European Organization for Nuclear Research) · 2025
This paper delves into the development of a novel real-time collaborative code editing application, fostering a harmonious environment for geographically dispersed development teams. The application, titled "Code Collab," empowers developers to work synchronously on the same codebase, eliminating the friction of traditional version control workflows. Built with robust technologies like [List technologies used in your project (e.g., programming languages, frameworks, libraries)], Code Collab prioritizes user experience and performance. The application offers a streamlined development environment with features akin to traditional IDEs, including syntax highlighting, code completion, and error checking. This familiarity empowers developers to transition seamlessly into a collaborative workflow. To facilitate real-time collaboration, the application leverages a robust Operational Transformation (OT) algorithm, ensuring synchronized editing experience for all collaborators. This paper explores the technical considerations addressed during development, including scalability achieved through container orchestration with Kubernetes. Additionally, the paper discusses the potential impact of Collaborative Symphony on software development practices, including enhanced communication, improved developer productivity, and geographically flexible development teams. By outlining the functionalities, technical considerations, and future prospects, this paper positions Collaborative Symphony as a transformative force in the realm of collaborative development.
Cited by 0Open access
Ryan Yen, Jian Zhao, Daniel Vogel · 2024
We present an initial step towards building a system for programmers to edit code using free-form sketch annotations drawn directly onto editor and output windows. Using a working prototype system as a technical probe, an exploratory study (N = 6) examines how programmers sketch to annotate Python code to communicate edits for an AI model to perform. The results reveal personalized workflow strategies and how similar annotations vary in abstractness and intention across different scenarios and users.
Cited by 2
Mercado, Yadiel, Torres, Gabriel, Alvarez, Michael · Zenodo (CERN European Organization for Nuclear Research) · 2025
Cited by 0Open access
Raab, Felix, Wolff, Christian, Echtler, Florian · University of Regensburg Publication Server (University of Regensburg) · 2013
Despite widespread use of touch-enabled devices, the field of software development has only slowly adopted new interaction methods for available tools. In this paper, we present our research on RefactorPad, a code editor for editing and restructuring source code on touchscreens. Since entering and modifying code with on-screen keyboards is time-consuming, we have developed a set of gestures that take program syntax into account and support common maintenance tasks on devices such as tablets. This work presents three main contributions: 1) a test setup that enables researchers and participants to collaboratively walk through code examples in real-time; 2) the results of a user study on editing source code with both finger and pen gestures; 3) a list of operations and some design guidelines for creators of code editors or software development environments who wish to optimize their tools for touchscreens.
Cited by 0
Alessandro Del Sole · Apress eBooks · 2023
Visual Studio Code is not just another evolved text editor with syntax colorization and automatic indentation. Instead, it is a very powerful codefocused development environment expressly designed to make it easier to write web, mobile, and cloud applications using languages that are available to different development platforms.
Cited by 0
2008
We present Linked Editing, a novel, lightweight editorbased technique for managing duplicated source code. Linked Editing is implemented in a prototype editor called Codelink. We argue that the use of programming abstractions like functions and macros—the traditional solution to duplicated code—has inherent cognitive costs, leading programmers to chronically copy and paste code instead. Our user study compares functional abstraction with Linked Editing and shows that Linked Editing can give the benefits of abstraction with orders of magnitude decrease in programming time. 1.
Cited by 0
Haverbeke, Marijin · TIB KMO / FLOWWORKS GmbH · 2021
Cited by 0Open access
M. Toomim, Andrew Begel, Stuart L. Graham · 2005
We present Linked Editing, a novel, lightweight editor-based technique for managing duplicated source code. Linked Editing is implemented in a prototype editor called Codelink. We argue that the use of programming abstractions like functions and macros-the traditional solution to duplicated code-has inherent cognitive costs, leading programmers to chronically copy and paste code instead. Our user study compares functional abstraction with Linked Editing and shows that Linked Editing can give the benefits of abstraction with orders of magnitude decrease in programming time.
Cited by 173
Mercado, Yadiel, Torres, Gabriel, Alvarez, Michael · Zenodo (CERN European Organization for Nuclear Research) · 2025
Cited by 0Open access
Sampaio, Afonso Pinheiro · Repositório Institucional do ISCTE-IUL (ISCTE-IUL) · 2024
In contemporary software development, collaboration among developers is a fundamental aspect of the management and evolution of codebases. Version control systems such as Git facilitate collaboration by enabling the integration of version branches into a primary development branch. However, the occurrence of parallel modifications frequently gives rise to merge conflicts, which have the potential to disrupt workflows and delay development timelines. Although automated tools can address some conflicts, manual intervention is frequently necessary to resolve instances of overlapping changes. In order to prevent the accumulation of these conflicts, the practice of continuous integration has been proposed as a means of encouraging frequent code integration, with the aim of improving conflict management. This dissertation proposes a protocol for propagation of code changes to facilitate the integration of continuous integration practices directly into collaborative coding environments. This approach enables the early detection of conflicts and streamlines the merge process. The Javardair prototype was developed with the objective of validating this approach for Java, employing a transformation-based methodology. In contrast to conventional version control systems, this approach represents code modifications as structured, semantic-level transformations within an abstract syntax tree instead of a textual level. This facilitates a more precise identification of conflicts, as it is based on code semantics rather than simple line-by-line comparisons and also enables a more effective application of modifications.
Cited by 0Open access
Gagnon, Joshua A. · Digital WPI · 2006
SciTE is an open source text editor that supports syntax styling, folding, code completion, and call tips. This project delivered a graphical options pane to replace the old method of editing options and also added a debugging interface. These changes are the beginning of many possible enhancements to make SciTE a complete lightweight integrated development environment.
Cited by 0
Han, Sangmok · DSpace@MIT (Massachusetts Institute of Technology) · 2011
Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 2011.
Cited by 0Open access
Toru Suzuki, Maki Asami, Sanjay G. Patel et al. · Scientific Reports · 2018
Multiple applications of genome editing by CRISPR-Cas9 necessitate stringent regulation and Cas9 variants have accordingly been generated whose activity responds to small ligands, temperature or light. However, these approaches are often impracticable, for example in clinical therapeutic genome editing in situ or gene drives in which environmentally-compatible control is paramount. With this in mind, we have developed heritable Cas9-mediated mammalian genome editing that is acutely controlled by the cheap lysine derivative, Lys(Boc) (BOC). Genetic code expansion permitted non-physiological BOC incorporation such that Cas9 (Cas9<sup>BOC</sup>) was expressed in a full-length, active form in cultured somatic cells only after BOC exposure. Stringently BOC-dependent, heritable editing of transgenic and native genomic loci occurred when Cas9<sup>BOC</sup> was expressed at the onset of mouse embryonic development from cRNA or Cas9<sup>BOC</sup> transgenic females. The tightly controlled Cas9 editing system reported here promises to have broad applications and is a first step towards purposed, spatiotemporal gene drive regulation over large geographical ranges.
Cited by 18Open access
Roberto Cattaneo · Cellular and Molecular Life Sciences · 1990
Cited by 17Open access