{"id":253857,"date":"2025-08-14T05:00:59","date_gmt":"2025-08-14T12:00:59","guid":{"rendered":"https:\/\/devblogs.microsoft.com\/visualstudio\/?p=253857"},"modified":"2025-08-13T12:51:27","modified_gmt":"2025-08-13T19:51:27","slug":"improving-codebase-awareness-in-visual-studio-chat","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/visualstudio\/improving-codebase-awareness-in-visual-studio-chat\/","title":{"rendered":"Improving Codebase Awareness in Visual Studio Chat"},"content":{"rendered":"

Smarter Code Search in Visual Studio: From BM25 to Semantic Search<\/span><\/b>\u00a0<\/span><\/h2>\n

In our latest 17.14.11 release, we\u2019ve made a significant leap forward in how we explore your code to retrieve meaningful context. Our new Remote Semantic Search integration helps you find exactly what you need faster and with greater precision than ever before.<\/span>\u00a0<\/span><\/p>\n

By embedding Remote Semantic Search directly into the Visual Studio Copilot code search experience, we\u2019ve combined the power of traditional keyword search (BM25) with the deep contextual understanding of cutting-edge AI models. This means your searches go beyond just matching words, they grasp the concepts and intent behind your search queries.<\/span>\u00a0<\/span><\/p>\n

Let\u2019s dive into what\u2019s new, why it\u2019s a game-changer, and how you can start using it today.<\/span>\u00a0<\/span><\/p>\n

The Legacy: BM25 + Reranking<\/span><\/b>\u00a0<\/span><\/h3>\n

Before this update, Visual Studio\u2019s code search relied solely on a BM25-based search engine. BM25 is a well-established ranking algorithm that evaluates how closely each document, such as code files or symbols, matches your query. It does this by analyzing term frequency and document length, balancing how often a term appears within a document against how rare it is across the entire codebase. Simply put, the more a term shows up in a file (to a certain extent) and the less common it is overall, the more relevant that file is considered.<\/span>\u00a0<\/span><\/p>\n

On top of that, we used reranking logic to fine-tune results. This step gave extra weight to certain matches, like hits in file names or symbols within active projects, so that the most likely relevant results would surface right at the top.<\/span>\u00a0<\/span><\/p>\n

The limitations:<\/span><\/b>\u00a0<\/span><\/h4>\n