Modern recommendation systems can be broadly divided into two key stages: the ranking stage, where the system predicts various user engagements (e.g., click-through rate, like rate, follow rate, watch time), and the value model stage, which aggregates these predictive scores through a function (e.g., a linear combination defined by a weight vector) to measure the value of each content by a single numerical score. Both stages play roughly equally important roles in real industrial systems; however, how to optimize the model weights for the second stage still lacks systematic study. This paper focuses on optimizing the second stage through auto-tuning technology. Although general auto-tuning systems and solutions - both from established production practices and open-source solutions - can address this problem, they typically require weeks or even months to identify a feasible solution. Such prolonged tuning processes are unacceptable in production environments for recommendation systems, as suboptimal value models can severely degrade user experience. An effective auto-tuning solution is required to identify a viable model within 2-3 days, rather than the extended timelines typically associated with existing approaches. In this paper, we introduce a practical auto-tuning system named HyperZero that addresses these time constraints while effectively solving the unique challenges inherent in modern recommendation systems. Moreover, this framework has the potential to be expanded to broader tuning tasks within recommendation systems.

Embedding Based Retrieval (EBR) is a crucial component of the retrieval stage in (Ads) Recommendation System that utilizes Two Tower or Siamese Networks to learn embeddings for both users and items (ads). It then employs an Approximate Nearest Neighbor Search (ANN) to efficiently retrieve the most relevant ads for a specific user. Despite the recent rise to popularity in the industry, they have a couple of limitations. Firstly, Two Tower model architecture uses a single dot product interaction which despite their efficiency fail to capture the data distribution in practice. Secondly, the centroid representation and cluster assignment, which are components of ANN, occur after the training process has been completed. As a result, they do not take into account the optimization criteria used for retrieval model. In this paper, we present Hierarchical Structured Neural Network (HSNN), a deployed jointly optimized hierarchical clustering and neural network model that can take advantage of sophisticated interactions and model architectures that are more common in the ranking stages while maintaining a sub-linear inference cost. We achieve 6.5% improvement in offline evaluation and also demonstrate 1.22% online gains through A/B experiments. HSNN has been successfully deployed into the Ads Recommendation system and is currently handling major portion of the traffic. The paper shares our experience in developing this system, dealing with challenges like freshness, volatility, cold start recommendations, cluster collapse and lessons deploying the model in a large scale retrieval production system.

The rise of large language models (LLMs) has unlocked various applications of this technology in software development. In particular, generative LLMs have been shown to effectively power AI-based code authoring tools that can suggest entire statements or blocks of code during code authoring. In this paper we present CodeCompose, an AI-assisted code authoring tool developed and deployed at Meta internally. CodeCompose is based on the InCoder LLM that merges generative capabilities with bi-directionality. We have scaled up CodeCompose to serve tens of thousands of developers at Meta, across 10+ programming languages and several coding surfaces. We discuss unique challenges in terms of user experience and metrics that arise when deploying such tools in large-scale industrial settings. We present our experience in making design decisions about the model and system architecture for CodeCompose that addresses these challenges. Finally, we present metrics from our large-scale deployment of CodeCompose that shows its impact on Meta's internal code authoring experience over a 15-day time window, where 4.5 million suggestions were made by CodeCompose. Quantitative metrics reveal that (i) CodeCompose has an acceptance rate of 22% across several languages, and (ii) 8% of the code typed by users of CodeCompose is through accepting code suggestions from CodeCompose. Qualitative feedback indicates an overwhelming 91.5% positive reception for CodeCompose. In addition to assisting with code authoring, CodeCompose is also introducing other positive side effects such as encouraging developers to generate more in-code documentation, helping them with the discovery of new APIs, etc.

These companies are leading the way in artificial intelligence. Artificial intelligence was once a far-off imagination of scientists and sci-fi enthusiasts. Now, the industry has a value just under $1 trillion and is projected to grow to $14 trillion by 2030, according to Ark Invest. That's because AI is more than just a supercomputer that can play chess and engage in small talk; companies are using AI to automate and streamline their business processes. For example, automated algorithms remove most of the posts that violate Facebook's community standards.

Meta Platforms Inc. will change its ad delivery system to address concerns that it violates the Fair Housing Act by discriminating against users, as part of a settlement with a federal regulator. The accord resolves a lawsuit by the US Department of Housing and Urban Development alleging that the algorithms used in Meta's advertising systems allowed marketers to violate fair housing laws by limiting or blocking certain groups of people from seeing housing ads on the service. "Because of this ground-breaking lawsuit, Meta will--for the first time--change its ad delivery system to address algorithmic discrimination," Manhattan US Attorney Damian Williams said in a statement. Meta said Tuesday that it built machine learning technology to ensure that ads reach people that reflect the overall potential audience for a particular ad, and not just a subset of that group. In a blog post, Meta wrote that it will "work to ensure the age, gender and estimated race or ethnicity of a housing ad's overall audience matches the age, gender, and estimated race or ethnicity mix of the population eligible to see that ad."

Thinking artificial intelligence could boost your portfolio right about now? But which AI stocks beyond the usual suspects: Alphabet Inc. (NASDAQ: GOOG), Apple Inc. (NASDAQ: AAPL), Twilio Inc. (NYSE: TWLO), ServiceNow, Inc. (NYSE: NOW), NVIDIA Corporation (NASDAQ: NVDA) and QUALCOMM Incorporated (NASDAQ: QCOM). Whether you believe some of the famous companies above are bearish and/or overvalued (another argument for another day) or are just looking to inject some new blood into your portfolio, let's go through some options you might consider. But first, in the interest of education, what are AI stocks and which ones should you consider right now? Let's find out. At its most basic level, artificial intelligence (AI) refers to an algorithm or dynamic machine that learns and interprets the data given to it.

AI stocks may be excellent long-term investments. The global artificial intelligence market is on track to hit more than $554 billion in total revenue by 2024, according to market researcher IDC. AI, automation and robotics are disrupting virtually every major industry. From machine learning to the "internet of things," autonomous vehicles, virtual assistants and smart home appliances, companies that aren't embracing AI and incorporating it into their business models risk going obsolete. Countless companies stand to benefit from AI, but a handful of companies have business models focused specifically on automation.