# Meta AI Research Engineering at Meta Blog This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## An Organizational Second Brain: Building an AI That Learns From Experts DevFeed: [An Organizational Second Brain: Building an AI That Learns From Experts]() Original publisher: [Read original article]() Author: Shaurya Sengar; Jason Nawrocki; Jay Shah; Prashant Kommireddi Published: 2026-09-02T09:00:29Z Content type: article Language: en Sources: [Engineering at Meta](), [Meta AI Research](), [Meta ML Applications]() Topics: [Large Language Model](), [Architecture & Design]() Tags: [ai-agent](), [ai-research](), [architecture](), [compliance](), [llms](), [ml-applications](), [security-privacy]() ### AI overview Meta describes an AI agent for a compliance domain that preserves specialist knowledge through an auditable knowledge architecture, an expert-like reasoning layer, and a feedback-driven improvement pipeline without model retraining. ### Source excerpt We've built an AI agent that acts as a secondary expert for a given domain, making deep specialist knowledge readily available and preserved for anyone in an organization to access, share, and build upon. This is not a typical domain-specific agent. Its novelty comes from integrating two layers: A structured, auditable knowledge architecture separates what [...] Read More... The post An Organizational Second Brain: Building an AI That Learns From Experts appeared first on Engineering at Meta. ## GEM Training: How Meta Doubled the Efficiency of Its LLM-Scale Ads Foundation Model DevFeed: [GEM Training: How Meta Doubled the Efficiency of Its LLM-Scale Ads Foundation Model]() Original publisher: [Read original article]() Author: Darren Liu; Huayu Li; Raghav Boinepalli; Yuzhen Huang; Jackie (Jiaqi) Xu; Richard Qiu; Chunzhi Yang; Rich Zhu; Dev (Devashish) Shankar; Huaqing Xiong Published: 2026-08-03T18:00:17Z Content type: article Language: en Sources: [Engineering at Meta](), [Meta AI Research](), [Meta ML Applications]() Topics: [recommendation systems](), [networking]() Tags: [ads](), [ai-infrastructure](), [ai-research](), [data-infrastructure](), [gpu](), [hardware](), [kernels](), [llm](), [meta](), [ml-applications](), [networking](), [recommendation-systems](), [training]() ### AI overview Meta describes training its GEM ads recommendation foundation model at LLM scale. The article covers recommendation-specific kernels, ultra-low-precision training, and topology-aware parallelism that doubled end-to-end training efficiency to 20-25% MFU while increasing training FLOPs fourfold. ### Source excerpt Meta's Generative Ads Recommendation Model (GEM), the foundation model behind ads recommendations across Instagram and Facebook, now trains at LLM scale on several thousand of the latest-generation GPUs. This post goes into the details on how we achieved: doubling end-to-end (E2E) training efficiency to 20-25% Model FLOPs Utilization (MFU) while scaling training FLOPs 4x in [...] Read More... The post GEM Training: How Meta Doubled the Efficiency of Its LLM-Scale Ads Foundation Model appeared first on Engineering at Meta. ## Exploring Hierarchical Interest Representation For Meta Ads Deep Funnel Optimization DevFeed: [Exploring Hierarchical Interest Representation For Meta Ads Deep Funnel Optimization]() Original publisher: [Read original article]() Author: Yuhui Ouyang; Di Wang; Sreedal Menon; Jie Tian Published: 2026-07-15T17:00:52Z Content type: article Language: en Sources: [Engineering at Meta](), [Meta AI Research]() Topics: [Optimization](), [Embeddings](), [Graphs](), [recommendation systems](), [LLMs](), [Artificial Intelligence](), [Transformer](), [data]() Tags: [ads](), [ai](), [ai-research](), [data](), [embeddings](), [engineering](), [frontier-ai](), [generative](), [learning](), [meta](), [optimization](), [personalization](), [reasoning](), [research](), [retrieval](), [scale]() ### AI overview Meta describes Hierarchical Interest Representation, an upstream system that learns unified embeddings for users, advertisers, products, and services. It combines graph learning, multimodal content processed through LLMs, engagement signals, and self-supervised distillation to improve personalization, retrieval, ranking, and deep-funnel advertising optimization. ### Source excerpt Hierarchical Interest Representation is a research area for Meta Ads. We're exploring an upstream representation layer over the universe of Ads entities - users, advertisers, products, services - learning unified embeddings that connect users' inferred interests with the breadth of what advertisers offer in their deep funnel ads. The innovations in Hierarchical Interest Representation are [...] Read More... The post Exploring Hierarchical Interest Representation For Meta Ads Deep Funnel Optimization appeared first on Engineering at Meta. ## 10 Years of Meta's Commitment to Python DevFeed: [10 Years of Meta's Commitment to Python]() Original publisher: [Read original article]() Author: Chris Wiltz Published: 2026-06-30T16:00:46Z Content type: opinion Language: en Sources: [Meta AI Research](), [Meta ML Applications]() Topics: [Meta](), [Python](), [Open Source](), [Programming language](), [Software](), [PyTorch]() Tags: [ai-research](), [culture](), [data-infrastructure](), [devinfra](), [engineering](), [meta](), [ml-applications](), [open-source](), [production-engineering](), [programming](), [python](), [pytorch](), [software]() ### AI overview Meta reflects on its 10th consecutive year sponsoring the Python Software Foundation and explains Python's importance across its engineering stack, including products, infrastructure, and AI research. ### Source excerpt This year marks Meta's 10th consecutive year as a sponsor of the Python Software Foundation (PSF), the charitable organization dedicated to advancing, supporting, and protecting the open-source Python programming language and the community that sustains it. Python is one of the world's most influential programming languages, and we use it across our engineering stack, from [...] Read More... The post 10 Years of Meta's Commitment to Python appeared first on Engineering at Meta. ## RCCLX: Innovating GPU Communications on AMD Platforms DevFeed: [RCCLX: Innovating GPU Communications on AMD Platforms]() Original publisher: [Read original article]() Author: Sudharssun Subramanian; Subodh Iyengar; Cen Zhao; Srinath Bayareddy; James Hongyi Zeng Published: 2026-02-24T21:30:54Z Content type: article Language: en Sources: [Meta AI Research]() Topics: [GPU](), [Meta](), [communications](), [AI Models](), [Latency](), [Inference](), [Large Language Model](), [Transformer]() Tags: [accelerate](), [ai](), [ai-models](), [ai-research](), [communications](), [data-center-engineering](), [gpu](), [inference](), [large-language-model](), [latency](), [layer](), [meta](), [ml-applications](), [networking-traffic](), [open-source](), [parallelism]() ### AI overview Meta describes the initial open-source release of RCCLX, an enhanced version of RCCL for AMD platforms integrated with Torchcomms. The article presents Direct Data Access algorithms and Low Precision Collectives, including approaches intended to reduce communication latency during large language model inference. ### Source excerpt We are open-sourcing the initial version of RCCLX - an enhanced version of RCCL that we developed and tested on Meta's internal workloads. RCCLX is fully integrated with Torchcomms and aims to empower researchers and developers to accelerate innovation, regardless of their chosen backend. Communication patterns for AI models are constantly evolving, as are hardware [...] Read More... The post RCCLX: Innovating GPU Communications on AMD Platforms appeared first on Engineering at Meta. ## Scaling LLM Inference: Innovations in Tensor Parallelism, Context Parallelism, and Expert Parallelism DevFeed: [Scaling LLM Inference: Innovations in Tensor Parallelism, Context Parallelism, and Expert Parallelism]() Original publisher: [Read original article]() Author: Cen Zhao; Xiaodong Wang; Jianyu Huang Published: 2025-10-17T16:00:50Z Content type: article Language: en Sources: [Meta AI Research]() Topics: [Large Language Model](), [AI Inference](), [Inference Performance](), [LLM Techniques](), [GPU](), [sharding](), [long-context]() Tags: [ai-research](), [gpu](), [inference](), [kv-cache](), [large-language-models-llms](), [latency](), [llms](), [parallelism](), [performance](), [sharding]() ### AI overview Meta describes three forms of parallelism--tensor, context, and expert parallelism--for scaling large language model inference across GPUs. The article explains how prefill and decoding differ computationally and how these techniques target resource efficiency, throughput, and latency. ### Source excerpt At Meta, we are constantly pushing the boundaries of LLM inference systems to power applications such as the Meta AI App. We're sharing how we developed and implemented advanced parallelism techniques to optimize key performance metrics related to resource efficiency, throughput, and latency. The rapid evolution of large language models (LLMs) has ushered in a [...] Read More... The post Scaling LLM Inference: Innovations in Tensor Parallelism, Context Parallelism, and Expert Parallelism appeared first on Engineering at Meta. ## Meta's ACH Tool Uses LLMs for Mutation-Guided Test Generation and Compliance Testing DevFeed: [Meta's ACH Tool Uses LLMs for Mutation-Guided Test Generation and Compliance Testing]() Original publisher: [Read original article]() Author: Mark Harman Published: 2025-09-30T16:00:08Z Content type: article Language: en Sources: [Meta AI Research]() Topics: [Testing](), [mutation-testing](), [Large Language Model](), [Meta](), [Artificial Intelligence](), [Software]() Tags: [ai](), [ai-research](), [code](), [llms](), [meta](), [ml-applications](), [mutation-testing](), [security-privacy](), [testing]() ### AI overview Meta describes its Automated Compliance Hardening (ACH) tool, which uses large language models to generate relevant code mutants and tests designed to catch them. The article explains how this supports mutation testing and helps identify compliance-related bugs. ### Source excerpt Following our keynote presentations at FSE 2025 and Eurostar 2025, we're delving further into the development of Meta's Automated Compliance Hardening (ACH) tool, an LLM-based tool for software testing that is automating aspects of compliance adherence at Meta, while accelerating developer and product velocity. By leveraging LLMs we've been able to overcome the barriers that [...] Read More... The post LLMs Are the Key to Mutation Testing and Better Compliance appeared first on Engineering at Meta. ## Meta 3D AssetGen: Generating 3D Worlds With AI DevFeed: [Meta 3D AssetGen: Generating 3D Worlds With AI]() Original publisher: [Read original article]() Author: Pascal Hartig Published: 2025-09-29T14:00:42Z Content type: article Language: en Sources: [Meta AI Research]() Topics: [3D](), [Artificial Intelligence](), [foundation-models](), [Large Language Model](), [Meta]() Tags: [3d](), [ai](), [ai-research](), [foundation](), [images](), [keynote](), [learning](), [llms](), [meta](), [meta-tech-podcast](), [ml-applications](), [model](), [podcast](), [text](), [virtual](), [virtual-reality](), [worlds]() ### AI overview A Meta Tech Podcast episode discusses AssetGen, a foundation model for 3D assets. Meta's XR Tech team explains how it was built and trained, the role of LLMs in VR, and efforts to generate complete 3D worlds from text prompts. ### Source excerpt Imagine being able to use AI to create 3D virtual worlds using prompts as easily as you can generate images. The intersection of AI and VR was one of the biggest topics at Meta Connect this year. In his keynote, Mark Zuckerberg shared his vision of a future where anyone can create virtual worlds using [...] Read More... The post Meta 3D AssetGen: Generating 3D Worlds With AI appeared first on Engineering at Meta. ## Meta's Infrastructure Evolution and the Advent of AI DevFeed: [Meta's Infrastructure Evolution and the Advent of AI]() Original publisher: [Read original article]() Author: Yee Jiun Song; Kaushik Veeraraghavan Published: 2025-09-29T13:00:15Z Content type: article Language: en Sources: [Meta AI Research]() Topics: [Meta](), [Artificial Intelligence](), [Hardware](), [Network](), [cloud-infrastructure](), [Terraform]() Tags: [ai](), [ai-research](), [apache](), [data-center-engineering](), [data-centers](), [data-infrastructure](), [data-management](), [database](), [devinfra](), [hardware](), [infrastructure](), [lamp](), [linux](), [meta](), [ml-applications](), [mysql](), [network](), [networking-traffic](), [open-source](), [production-engineering](), [scale]() ### AI overview Meta describes how its infrastructure evolved from a small university-focused social network into a globally networked operation serving more than 3.4 billion people. The article explains that AI has changed infrastructure-scaling assumptions and requires innovation across hardware, software, networks, and data centers, while outlining earlier database, caching, social graph, ranking, and photo-service scaling work. ### Source excerpt Over the past 21 years, Meta has grown exponentially from a small social network connecting a few thousand people in a handful of universities in the U.S. into several apps and novel hardware products that serve over 3.4 billion people throughout the world. Our infrastructure has evolved significantly over the years, growing from a [...] Read More... The post Meta's Infrastructure Evolution and the Advent of AI appeared first on Engineering at Meta.