Microsoft Research Blog

English

  1. Beyond Accuracy: Realistic and Diagnostic Evaluation of Code Generation Models 

    November 25, 2025 | Pareesa Ameneh Golnari, Xiaoyu Liu (lixiaoyu), Gabriel Ryan (ryangabriel), and Shengyu Fu (shengyfu)

    DevBench is a telemetry-driven benchmark designed to evaluate Large Language Models (LLMs) on realistic code completion tasks. It includes 1,800 evaluation instances across six programming languages and six task categories derived from real developer telemetry, such as API usage and code purpose understanding. Unlike prior…

  2. Workload Intelligence: Workload-Aware IaaS Abstraction for Cloud Efficiency 

    November 25, 2025

    Today, cloud workloads are largely opaque to the cloud platform. Typically, the only information the platform receives is the virtual machine (VM) type and possibly a decoration to the type (e.g., the VM is evictable). Similarly, workloads receive minimal information from the platform; generally, only…

  3. Spatial Audio Rendering for Speech Live Translation 

    November 24, 2025 | Margarita Geleta

    Language barriers in virtual meetings remain a persistent challenge to global collaboration. While real-time translation technologies offer a promising solution, their integration into conversational interfaces often neglects key perceptual cues. This study explores how spatial audio rendering of translated speech affects comprehension, cognitive load, and…

  4. Three white line icons on a blue-to-green gradient background: a computer monitor with a globe symbol on the left, a cursor arrow with click lines in the center, and a computer mouse outline on the right.

    Fara-7B: An Efficient Agentic Model for Computer Use 

    November 24, 2025

    Fara-7B is our first agentic small language model for computer use. This experimental model includes robust safety measures to aid responsible deployment. Despite its size, Fara-7B holds its own against larger, more resource-intensive agentic systems.

  5. Cambridge Internship Program – Network Systems for Physical AI 

    November 24, 2025

    Drive innovation at Microsoft Research Cambridge by joining a world-class team pioneering breakthroughs in robotics and AI infrastructure. Collaborate across disciplines to develop impactful solutions for datacenter automation and next-generation intelligent systems. Work alongside leading experts, leverage state-of-the-art technology, and contribute to projects that shape…

  6. Fara-7B: An Efficient Agentic Model for Computer Use 

    November 24, 2025

    Progress in computer use agents (CUAs) has been constrained by the absence of large and high-quality datasets that capture how humans interact with a computer. While LLMs have thrived on abundant textual data, no comparable corpus exists for CUA trajectories. To address these gaps, we…

  7. CHERI-Lite for Memory Safety Exploit Mitigation 

    November 24, 2025 | Tony Chen

    This paper proposes adopting the CHERI concept of tagging pointers and only allowing tagged pointers to be used to specify the address of load, store, and instruction-fetch operations. However, we propose keeping the pointers at 64 bits and thus, need to forgo the bounds checking…

  12. Senior Researcher – Generative AI – Microsoft Research AI Frontiers 

    November 22, 2025

    The mission of the AI Frontiers lab is to expand the pareto frontier of AI capabilities, efficiency, and safety through innovations in foundation models and learning agent platforms. We are seeking a Senior Researcher - Generative AI to join our team in Redmond, WA or New…