Explore Projects

Pwndbg delivers a powerful debugging experience on traditional Linux systems, but its support for Android environments remains limited. This is largely due to fundamental differences between GNU libc and Android’s Bionic, as well as the lack of tooling for Android-specific components such as Binder. As a result, many existing features either do not work as expected or provide limited insight when analyzing Android binaries. This project aims to address these limitations by introducing structured and practical support for Android within Pwndbg. It will extend existing features to correctly handle Bionic internals, particularly in memory inspection and heap analysis, and improve Binder support by enabling meaningful userspace inspection. Additionally, the project will refine internal abstractions to ensure consistent behavior across LLDB and GDB.

<p>project involve adding new transformations to spirv-fuzz tools to improve their support for compute shaders</p>

LiquidHaskell is a verification tool that allows Haskell developers to prove the correctness of their code through refinement types. This proposal aims to restore the support for type class elaboration: a feature that enables the verification of typeclasses laws and instances. The work that had been done previously to support typeclass elaboration is currently broken: typeclass methods produce "Unknown logic name" errors, hence the test suite for the feature has been disabled. The aim of this project is to restore this functionality, by fixing the elaboration mechanism responsible for making typeclass methods available during verification. I will trace the elaboration pipeline, implement the necessary fixes and extend test coverage. Additionally, the project will include an in-depth analysis of the features supported by the implementation and how they compare against the expected outcomes in the original paper. The deliverables will include successfully merged contributions on the LiquidHaskell GitHub repository, with its corresponding documentation, and a post with the solution on the Tweag blog.

The ENIGMA project uses a hierarchy of objects to represent entities. These objects are serialized and deserialized using (de)serialization routines implemented for them. However, the current implementation is not very clean as it checks the type of each argument against an if-else ladder, making it difficult to read and extend. This project aims to make the serialization framework easier to use and extend by changing the way the serialize, serialize_into and deserialize functions work to dispatch to supported types implicitly without the need for an if-else ladder. And the project also aims to implement support for additional data types like generic routines for map, vector and complex numbers. This would enable more easily serializing ENIGMA's internal state for use in sending state either across a network or into a file. Additionally, the project aims to introduce support for the JSON format within the serialization system, offering it as a new option alongside the existing bytes format. This addition enhances the readability and debugging capability of the serialized data.

This project aims to enhance the BEAD framework for anomaly detection in high-energy physics by developing a new multi-stage architecture that combines a contrastive variational autoencoder (VAE) with a self-supervised transformer and multi-task fine-tuning layers. The key goal is to learn generator-invariant representations from 4-momentum vectors of background events by pretraining the VAE using generator labels (e.g., Pythia, Herwig, Sherpa). The latent variables produced are then masked and enriched with high-level physics features (e.g., MET, HT, Meff) and passed through a transformer trained to reconstruct them, encouraging deeper physics representation learning. The enhanced latent space is decoded using fine-tuning layers that perform multiple tasks simultaneously: reconstructing event-level features, predicting jet properties, and regressing unseen quantities. This structure encourages collaboration between the VAE and transformer components, increasing robustness against spurious correlations and improving anomaly detection performance. Deliverables will include: - A modular implementation of the full architecture within BEAD - Integration with tools like Weights & Biases for experiment tracking - Docker support for reproducibility - Evaluation benchmarks on synthetic datasets - Final documentation and presentation of findings at the CERN ML Forum

Computers with ARM architecture are prevalent but GNU Boot currently doesn't support them. My focus has also shifted from the classic x86_64 laptops compatible with fully free BIOS to newer armv7 and beyond to armv8 computers. The defunct fully free Libreboot has supported a few Chromebooks that didn't require blobs to boot, thanks to efforts of contributors such as PaulK. I checked with him for basic guidance and I'm in need of a mentor from the GNU Boot project under the GNU umbrella (as GSoC 2026 approved organization). Having GNU Boot support ARM computers would play well into the efforts of fully free GNU/Linux-libre distributions such as Trisquel to complete freedom-respecting ARM laptops. My work will involve updating the previous code in Libreboot and (PaulK's fork libettereboot) for the Chromebook C201 and C100 laptops and porting GNU Boot to newer ARM targets found in coreboot. Since I'm a Embedded Systems Master's student, Google Summer of Code seems to be the perfect opportunity to extend the GNU Boot contributor team while providing basic funding for taking GNU Boot to the next architecture.

<p>The idea is to implement cloud support in ScummVM. Users would be able to sync their save files and download game data from different cloud storage services on all their devices.</p> <p>At least these cloud services would be supported:</p> <ul> <li>Dropbox;</li> <li>Box;</li> <li>OneDrive;</li> <li>Google Drive.</li> </ul> <p>All of them use OAuth2 and provide an API to manage user's files.</p> <p>Things to be implemented:</p> <ul> <li>GUI for user to interact with;</li> <li>more or less portable way to open a link in a browser (for OAuth);</li> <li>more or less portable way to run a local webserver (for OAuth);</li> <li>portable (using libcurl) API interaction for all supported cloud services.</li> </ul> <p>Possible mentors: <a href="http://scummvm.org/index.php/User:Sev" target="_blank">Eugene Sandulenko</a>, <a href="http://scummvm.org/index.php/User:Uruk" target="_blank">Peter Bozso</a>.</p>

<p>KDE Connect has recently launched an SMS Messaging app for the desktop which uses KDE Connect to synchronize all the existing conversations from the remote device. In its current state it can:- Show a list of existing conversations and the corresponding conversation history. Send and display SMS. Send SMS to only existing contacts and conversations on the remote device. Display (not send) group messages. However, It is very limited to MMS and can only support showing text. It cannot send and display pictures, audio, video and other kinds of attachments, also users cannot reply to multi-target conversations. Considering the importance of sharing multimedia files and group messages, the SMS client becomes very limited in use. This project aims to improve MMS support with the integration of sending multi-target MMS by re-working the current implementation of sending SMS and UI support using Qt Multimedia library. By the end of this project, SMS client will be capable of sending and receiving MMS, will be able to send multi-target MMS with the capability to view pictures, video/audio in SMS app and the capability to download attachments.</p>

According to the Google Summer of Code (GSoC) project ideas listed for One proposed DeepChem project involves adding support for the OLMo Large Language Model (LLM) family to the DeepChem repository using the Hugging Face Model wrapper. The goal of this project is to enable DeepChem users to perform tasks such as text generation, classification, and continued pretraining with large language models, including Open Language Model (OLMo). I have already contributed several improvements to DeepChem’s HuggingFaceModel integration, including improving typing and validation for the fill mask method and implementing a new generate() function that wraps HuggingFace’s text generation API. I have also already demonstrated partial compatibility with OLMo models through recent contributions to the generate() function and example scripts using allenai/OLMo-1B-hf. I have also added support for causal language modeling in the HuggingFaceModel wrapper, including integration with AutoModelForCausalLM and end-to-end validation through unit tests. My contributions demonstrate familiarity with the DeepChem codebase and provide the foundation for extending DeepChem to support large language models such as OLMo.

<p>Graal produces analysis related to code complexity, quality, dependencies, vulnerability and licensing and the data produced conforms to the ones that can be processed by GrimoireLab. I will mainly be focusing on:</p> <ul> <li>Adding support of source code related metrics to Grimoirelab with the help of analysis data produced by Graal. </li> <li>Adapting Grimoirelab toolchain to be able to execute Graal and process the data produced by it.</li> <li>Writing appropriate unit tests for additional backends, their corresponding supporting connectors, and methods.</li> <li>Producing analytics related to proposed and calculated metrics* ( described below )</li> <li>Adding documentation related to additional features and improvements in existing ones.</li> </ul> <p>Out of all the five backends provided by Graal, CoCom (Code Complexity) covers a vast majority of the popular languages and CoLic (Code License) supported by NOMOS &amp; ScanCode helps us fetch license &amp; copyright related information from software development repositories and is language independent. Addition of metrics related to these two backends during GSoC period could be applied to a wide range of projects in the future.</p>

This project focuses on extending the imputeTestbench R package to better support modern time series data. The current package works well for univariate data, but many real-world datasets are multivariate, where multiple variables evolve together over time and may have structured missing values. The main goal is to add support for multivariate time series while keeping the package simple, stable, and fully compatible with CRAN. The project will introduce methods to handle multivariate inputs, simulate realistic missing data patterns such as block missingness, and improve benchmarking so that imputation methods can be evaluated both per variable and in aggregated form. In addition, the project will improve performance through efficient data handling and optional parallel processing using CRAN-safe approaches. The internal structure of the package will also be refined to improve readability, maintainability, and future extensibility. The expected deliverables include multivariate input support, structured missing data generation, improved benchmarking functions, optional parallel execution, and updated testing and documentation. These improvements will make imputeTestbench more useful for real-world applications such as environmental data, energy systems, finance, and sensor-based monitoring.

<h3>Summary</h3> <p>This project aims to implement Node Labeling and Preemption API, which are existing feature of resource managers, on REEF. Node Labels allow users to group resources with similar characteristics or with similar ACL policies and Preemption is a way to respect elasticity and SLAs together.</p> <h3>The Project</h3> <p>Even though YARN and Mesos, the resource managers have existing features that supports Node Labeling and Preemption, REEF doesn’t have APIs that makes these features available. To make developers can exploit these features on REEF regardless of which resource manager they use, APIs that make available Node Labeling and Preemption will be implemented on REEF in this project.</p> <h3>Required Deliverables</h3> <ul> <li>Support for the Node Labeling API as it exists in YARN and Mesos in REEF</li> <li>Support for the Preemption API as it exists in YARN and Mesos in REEF</li> </ul> <h3>Benefits for REEF</h3> <p>When Node Labeling and Preemption has been done in REEF, developer can exploit these features without directly dealing with resource managers such as YARN or Mesos, but through the same interface on REEF. It will increase portability and simplifies development of applications for cluster resource managers.</p>

<p><strong>TensorFlow Lite Flutter Support Suite</strong> is a cluster of three projects TfLite Flutter Plugin, TfLite Flutter Helper library, Flutter example ML app along with a tutorial.</p> <p><strong>TfLite Flutter Plugin</strong> provides TfLite JAVA like dart API to perform inference in flutter apps. The plugin is built on the top of the existing dart bindings and binds to TfLite C API using dart:ffi.</p> <p>View <a href="https://github.com/am15h/tflite_flutter_plugin/" target="_blank">am15h/tflite_flutter_plugin</a> &amp; <a href="https://pub.dev/packages/tflite_flutter" target="_blank">tflite_flutter on Pub</a></p> <p><strong>TfLite Flutter Helper library</strong> aims at providing an interface to easily use high-level objects involved in Image, Text, Audio processing with TfLite Flutter plugin, and related utility functions in addition to similar functionality as the TensorFlow Lite Android Support Library.</p> <p>View <a href="https://github.com/am15h/tflite_flutter_helper/" target="_blank">am15h/tflite_flutter_helper</a> &amp; <a href="https://pub.dev/packages/tflite_flutter_helper" target="_blank">tflite_flutter_helper on Pub</a></p> <p>A comprehensive and detailed <strong>Flutter example ML app</strong> demonstrating the combined efficient usage of the TfLite Flutter plugin and TfLite Flutter Helper package together will be published with a detailed tutorial.</p> <p>View <a href="https://github.com/am15h/object_detection_flutter/" target="_blank">am15h/object_detection_flutter</a> &amp; <a href="https://medium.com/@am15hg/real-time-object-detection-using-new-tensorflow-lite-flutter-support-ea41263e801d" target="_blank">blog</a></p>

<p>To support colored glyphs in <strong>OpenType</strong> fonts, several formats currently exist, namely <strong>SBIX</strong>, <strong>CBDT</strong> and <strong>COLR</strong>. These are used widely with varying degree of support across different platforms. <strong>SVG</strong> is a recent addition that uses a subset of <strong>SVG1.1</strong> features to represent glyphs. It is growing in popularity and will be supported across most devices very soon. <strong>FreeType</strong> is a very powerful font rendering library that is being used across billions of devices. It already supports <strong>SBIX</strong>, <strong>CBDT</strong> and <strong>COLR</strong> formats. The goal of this project is to add support for SVG glyphs to <strong>FreeType</strong>. I plan to accomplish this by:</p> <ol> <li>Writing code that reads SVG documents from SVG font tables and decode it if it's in <strong>gzip</strong> encoded format.</li> <li>Creating a callback API that would allow users of FreeType to plug any external SVG Renderer to be used for SVG rendering.</li> <li>Choose one SVG Renderer as the default one so that SVG support comes out of the box with FreeType, while making the callback API generic enough to support other renderers too. </li> <li>Testing and documenting everything.</li> </ol>

SQLancer is an open-source tool for testing the correctness of SQL database systems and supports close to 20 database systems. The goal of this project is to add support for one or more new database systems to SQLancer and test them to find potential bugs. As for which DBMS to add support for, I would like to firstly choose StoneDB, which is being actively maintained and their community showed their interest of being support by SQLancer. What’s more, StoneDB is just the primary choose. If we have time, we can also add support to other DBMS. The project Proposal will consist of four main stages: 1)Choose a DBMS 2)Implementation of selected DBMS 3)Refactoring existing code 4)Testing and reporting

Java Pathfinder (JPF) is a Java virtual machine that can run compiled Java programs as well as model checking it. It is also an extensible software analysis framework for Java bytecode. As it has been created for about two decades, supporting new Java language features is of vital importance for its wide adoption. JPF community has taken steps to add Java 11 support, but there are still some unsolved issues. In this project, I plan to add better Java 11 support for JPF. This includes more comprehensive Java bytecode support, especially for invokedynamic, more JDK API support, and some bug fixes. It will fix all 13 failing tests JPF faces now on Java 11 and enable more programs to run on JPF.

Description: Kubernetes Validating Admission Policy Support Expected Outcome: Kyverno support for ValidatingAdmissionPolicy in one of the identified proposals.

SQLancer randomly generates test cases consisting of potentially hundreds or thousands of SQL statements. In order to efficiently identify bugs in these statements, automatic reduction is needed to systematically remove the bug inducing statements until a minimal bug-triggering test case is derived. To improve the efficiency and effectiveness of SQLancer for bug detection and resolution in various DBMSs, I propose to undertake the following tasks: 1. Implement delta-debugging algorithms for statement reduction in SQLancer to improve efficiency and speed. 2. Define parser-tree transforms to simplify SQL statements. Since SQL syntax varies among DBMSs, generic parser-tree transforms that can be applied across multiple DBMSs will be implemented, such as removing expression nodes or removing certain clauses from the statement. 3. Generalize statement reduction to additional DBMSs by adding bug reproducers to them. The deliverables for this project include: 1. A delta-debugging algorithm that can reduce the bug-inducing statement. 2. Generic parser-tree transforms that simplify SQL statements. 3. Support of statement reduction for more DBMSs.

Clad is a Clang library which allows automatic differentiation (AD) of functions. Clang provides the necessary facilities for code transformation. The AD library can differentiate non-trivial functions, to find a partial derivative for trivial cases. Newer C++ versions provide the constexpr and consteval specifiers, but currently Clad does not abide by them. The aim of this proposal is to ensure that Clad generated derivative functions follow the semantics of the original functions. Adding support will likely require a new interface for CladFunction's and changes in clad::differentiate and clad::gradient.

<p>Cuckoo Sandbox is a malware analysis platform which performs basic static file analysis to in-depth dynamic analysis of binaries. Even though macOS modules exist, they are not being properly used by the Cuckoo work system. The idea is to add a long term support for macOS binaries by setting up kernel-space extension by rewriting <em>xnumon</em>. As xnumon in itself is an up to date monitoring module, we can rewrite the kext to incorporate our custom logging and reporting methods. Apart from that macOS malware specific auxiliary modules are to be written in the later part of the project</p>

In database design, composite primary keys are often necessary for the partitioning and sharding of database tables. - Citus is a PostgreSQL extension that transforms PostgreSQL into a distributed database. - django-multitenant is a library by Citus which enables developers to build multi-tenant applications in Django. In Citus, composite primary keys are required. So, in order to use these tools with Django, the developers must work around the Django ORM. To make building multi-tenant apps easier, I propose adding composite primary key support to the Django ORM. https://gist.github.com/csirmazbendeguz/de1fdf88bf9df0dfba837f8e88c35df6

<p>Accelerate high-dimensional spatial genomic matrix alignment and multi-sample normalization routines for large-scale spatial transcriptomics datasets.</p><p>Mentee will benchmark memory utilization and implement parallelized matrix transformations using C++ extensions inside BiocParallel.</p><p><br></p><p><strong>Deliverables:</strong></p><ul><li>Parallel C++ kernel extensions for BiocParallel.</li><li>Benchmark suite comparing CPU/RAM throughput.</li></ul><p><br></p><p><strong>Applications closing date: 15-Nov-2025</strong></p>

Current LHCb calorimeter reconstruction uses rule-based clustering without exploiting timing or learned representations. The PicoCal upgrade will provide picosecond timing and longitudinal segmentation, creating an opportunity for ML-based reconstruction. I propose a space-time kernel transformer that embeds timing as a physics-informed spatial coordinate for energy reconstruction, benchmarked against the Cellular Automaton and Graph Clustering baselines across four incremental dataset stages. Deliverables include a complete end-to-end pipeline, trained models with timing ablation study, benchmark tables with physics and latency metrics, and a documented open-source PyTorch Geometric codebase for the LHCb group.