Events & Conferences

Bringing practical applications of quantum computing closer

Published

3 years ago

March 11, 2022

The Annual Conference on Quantum Information Processing (QIP) — the major conference in the quantum information field — was held this week, and Amazon Web Services is its sole diamond sponsor.

Two Caltech professors who are also members of the AWS Center for Quantum Computing — Amazon Scholar John Preskill and Fernando Brandão, the director of quantum applications for Amazon’s Quantum Computing group — are coauthors on six papers at QIP (see sidebar).

But one of those — “Foundations for learning from noisy quantum experiments” — originated within Amazon’s Quantum Computing group, as did another QIP paper, “A randomized quantum algorithm for statistical phase estimation”.

“A randomized quantum algorithm for statistical phase estimation” describes a new method for statistical phase estimation, which could be used to calculate the ground-state energy of a molecule simulated by a quantum computer, among other applications. The technique requires fewer quantum bits (or qubits) to represent the molecule than existing methods do, and it also makes do with fewer gate operations, or manipulations of the quantum system.

Related content

As the major quantum computing conference celebrates its anniversary, we ask the conference chair and the head of Amazon’s quantum computing program to take stock.

“Foundations for learning from noisy quantum experiments” considers the case of a black-box quantum system — such as a noisy quantum computer — and shows that, if the system permits a particular set of quantum operations to be performed on it, its internal relationships can be accurately characterized. This means that near-term quantum computers with noisy qubits — that is, quantum computers that don’t always do what they’re supposed to — can still perform useful computations, because their operators can determine how noise is affecting the computational results.

Quantum computers

Where a bit in a classical computer can represent either 1 or 0, a qubit can represent 1, 0, or a quantum superposition of both. Perform a measurement on a qubit, however, and it falls out of superposition: it assumes a definite value, either 1 or 0.

If a group of qubits are entangled, so that their quantum properties depend on each other, then they can all share a single superposition. That superposition can be described as a probability distribution over definite states of the qubit array — states in which each qubit is either a 1 or a 0.

The probability distribution, in turn, is defined by a wave function, which has all the properties that electromagnetic waves do. When a sequence of measurements causes all the qubits to snap into definite states, the wave function is said to collapse.

Related content

Researchers affiliated with Amazon Web Services’ Center for Quantum Computing are presenting their work this week at the Conference on Quantum Information Processing.

Quantum computation consists of applying a series of operations — called gates, on the model of logic gates in classical computers — to an array of entangled qubits. For instance, a Hadamard gate puts a single qubit into superposition; a swap gate swaps two qubits.

The operations modify the qubits’ wave function so that it encodes some mathematical problem. When the wave function collapses, the definite values of the qubits represent — with high probability — the solution to the problem.

But maintaining entanglement across large numbers of qubits long enough to perform a useful computation is extremely difficult. To date, the largest quantum computer to exhibit entanglement has about 30 qubits. And most current qubits are “noisy”, or error prone.

Both the Amazon papers at QIP have a range of applications, but they’re well suited to the problem of near-term quantum computation, on devices with either limited numbers of qubits or noisy qubits.

Phase estimation

In 1994, when the world first learned of Peter Shor’s quantum algorithm for factoring numbers, it seemed that quantum computers might be able to solve an important class of problems — NP-complete problems — exponentially faster than classical computers.

Related content

New method enables entanglement between vacancy centers tuned to different wavelengths of light.

Now, that seems unlikely. But one thing quantum computers definitely will do better than classical computers is simulate quantum systems.

For instance, simulations can help chemists, materials scientists, and drug developers better understand the molecules they’re working with. But accurate simulation requires the modeling of quantum interactions between atoms, which would be much more efficient on a quantum computer than it is on today’s classical computers.

Molecular simulation is the problem addressed in “A randomized quantum algorithm for statistical phase estimation”. The first author on the paper is Kianna Wan, a graduate student at Stanford University who was an intern at Amazon when the work was done. She’s joined by Mario Berta, a senior research scientist in the AWS Quantum Computing group, and Earl Campbell, who was also an Amazon senior research scientist at the time.

When a molecule is simulated on a quantum computer, the phase of the qubits’ wave function can be used to compute the molecule’s ground-state energy. But because measurements on the qubits cause the wave function to collapse, estimating the energy requires a series of measurements, which repeatedly sample the wave function’s probability distribution.

A new implementation (bottom) of a Hadamard test (top) for phase estimation. The implementation applies a sequence of randomly selected quantum operations called Pauli operators (colored circles) and Pauli rotations (colored squares) to a quantum computer’s qubits. This procedure is applied multiple times, with different sequences of operations, to estimate the phase of a quantum wave function.

The number of qubits required to represent a molecule on a quantum computer is proportional to the size of the molecule. But existing methods of phase estimation require ancillary qubits — perhaps ten times the number of qubits required to represent the molecule — to encode the Hamiltonian matrix that represents the molecule’s energy function.

The Amazon researchers’ method allows more direct measurement of the qubits, because it uses importance sampling to preferentially sample the molecule’s strongest atomic interactions — the ones that contribute most to its overall energy.

Related content

How an Amazon quantum computing scientist won the first-ever quantum chess tournament.

This approach could end up requiring more samples than existing approaches. But given how hard qubits are to realize, in the near term, representing a molecule with, say, 100 qubits, and sampling those qubits more frequently, may be preferable to representing the molecule with 1,000 qubits and requiring fewer samples.

Learning from quantum experiments

In “Foundations for learning from noisy quantum experiments”, the researchers — first author Hsin-Yuan Huang, a Caltech graduate student who was an Amazon intern at the time; Steve Flammia, a principal research scientist at Amazon; and John Preskill, who’s Huang’s thesis advisor — consider a black-box quantum system: the experimenter can perform operations on the system and make measurements but otherwise has no idea how the system is internally configured.

In fact, the experimenter doesn’t know what effect the operations have on the system, nor what the measurements are measuring! Nonetheless, the authors prove a theorem stating that, if there exist operations that, in principle, allow the physical system to explore the full quantum Hilbert space — such as Hadamard gates and Toffoli gates — then it is possible to accurately characterize the system, including its noise properties.

The theorem is general: it could be useful for physical research on quantum-mechanical phenomena as well as quantum computing. But it has a clear application in the case of near-term quantum computers with noisy qubits. An accurate characterization of a noisy quantum computer could enable operators to devise experiments that yield useful results even given a certain probability of error.

Related content

The noted physicist answers 3 questions about the challenges of quantum computing and why he’s excited to be part of a technology development project.

Huang, Flammia, and Preskill also describe a pair of specific applications of their theory. The first is the use of neural networks to learn the characteristics of a quantum system.

They don’t use neural networks in the conventional way, however. Instead of simply providing sample inputs and outputs and letting the network learn correspondences between the two, they use the separate layers of the network to model consecutive operations applied to the quantum system and their results.

Within that formalism, however, they can use existing machine learning algorithms — gradient descent and backpropagation — to train the network. In a forthcoming paper, they show that, so long as the noise of the quantum system is below some threshold, this approach will yield a rigorous model of the system.

They also consider the case in which the qubits of a quantum computer are so noisy that rigorously characterizing them is impossible. Even in that case, they show, it’s possible to characterize the system well enough that on some computations, it can still afford speedups relative to classical computers.

Source link

Events & Conferences

A New Ranking Framework for Better Notification Quality on Instagram

Published

4 days ago

September 2, 2025

Xian Sun

We’re sharing how Meta is applying machine learning (ML) and diversity algorithms to improve notification quality and user experience.
We’ve introduced a diversity-aware notification ranking framework to reduce uniformity and deliver a more varied and engaging mix of notifications.
This new framework reduces the volume of notifications and drives higher engagement rates through more diverse outreach.

Notifications are one of the most powerful tools for bringing people back to Instagram and enhancing engagement. Whether it’s a friend liking your photo, another close friend posting a story, or a suggestion for a reel you might enjoy, notifications help surface moments that matter in real time.

Instagram leverages machine learning (ML) models to decide who should get a notification, when to send it, and what content to include. These models are trained to optimize for user positive engagement such as click-through-rate (CTR) – the probability of a user clicking a notification – as well as other metrics like time spent.

However, while engagement-optimized models are effective at driving interactions, there’s a risk that they might overprioritize the product types and authors someone has previously engaged with. This can lead to overexposure to the same creators or the same product types while overlooking other valuable and diverse experiences.

This means people could miss out on content that would give them a more balanced, satisfying, and enriched experience. Over time, this can make notifications feel spammy and increase the likelihood that people will disable them altogether.

The real challenge lies in finding the right balance: How can we introduce meaningful diversity into the notification experience without sacrificing the personalization and relevance people on Instagram have come to expect?

To tackle this, we’ve introduced a diversity-aware notification ranking framework that helps deliver more diverse, better curated, and less repetitive notifications. This framework has significantly reduced daily notification volume while improving CTR. It also introduces several benefits:

The extensibility of incorporating customized soft penalty (demotion) logic for each dimension, enabling more adaptive and sophisticated diversity strategies.
The flexibility of tuning demotion strength across dimensions like content, author, and product type via adjustable weights.
The integration of balancing personalization and diversity, ensuring notifications remain both relevant and varied.

The Risks of Notifications without Diversity

The issue of overexposure in notifications often shows up in two major ways:

Overexposure to the same author: People might receive notifications that are mostly about the same friend. For example, if someone often interacts with content from a particular friend, the system may continue surfacing notifications from that person alone – ignoring other friends they also engage with. This can feel repetitive and one-dimensional, reducing the overall value of notifications.

Overexposure to the same product surface: People might mostly receive notifications from the same product surface such as Stories, even when Feed or Reels could provide value. For example, someone may be interested in both reel and story notifications but has recently interacted more often with stories. Because the system heavily prioritizes past engagement, it sends only story notifications, overlooking the person’s broader interests.

Introducing Instagram’s Diversity-Aware Notification Ranking Framework

Instagram’s diversity-aware notification ranking framework is designed to enhance the notification experience by balancing the predicted potential for user engagement with the need for content diversity. This framework introduces a diversity layer on top of the existing engagement ML models, applying multiplicative penalties to the candidate scores generated by these models, as figure1, below, shows.

The diversity layer evaluates each notification candidate’s similarity to recently sent notifications across multiple dimensions such as content, author, notification type, and product surface. It then applies carefully calibrated penalties—expressed as multiplicative demotion factors—to downrank candidates that are too similar or repetitive. The adjusted scores are used to re-rank the candidates, enabling the system to select notifications that maintain high engagement potential while introducing meaningful diversity. In the end, the quality bar selects the top-ranked candidate that passes both the ranking and diversity criteria.

Figure.1: Instagram’s diversity-aware ranking framework where the diversity layer sits on top of the existing modeling layer and penalizes notifications that are too similar to recently sent ones.

Mathematical Formulation

Within the diversity layer, we apply a multiplicative demotion factor to the base relevance score of each candidate. Given a notification candidate 𝑐, we compute its final score as the product of its base ranking score and a diversity demotion multiplier:

$\text{Score}(c) = R(c) \times D(c)$

where R(c) represents the candidate’s base relevance score, and D(c) ∈ [0,1] is a penalty factor that reduces the score based on similarity to recently sent notifications. We define a set of semantic dimensions (e.g., author, product type) along which we want to promote diversity. For each dimension i, we compute a similarity signal p_i(c) between candidate c and the set of historical notifications H, using a maximal marginal relevance (MMR) approach:

$p_i(c) = \mathrm{max}_{h \in H}\mathrm{sim}_i(c, h)$

where sim_i(·,·) is a predefined similarity function for dimension i. In our baseline implementation, p_i(c) is binary: it equals 1 if the similarity exceeds a threshold 𝜏_i and 0 otherwise.

The final demotion multiplier is defined as:

$D(c) = \prod_{i=1}^{m} \left( 1 - w_i \cdot p_i(c) \right)$

where each w_i∈ [0,1] controls the strength of demotion for its respective dimension. This formulation ensures that candidates similar to previously delivered notifications along one or more dimensions are proportionally down-weighted, reducing redundancy and promoting content variation. The use of a multiplicative penalty allows for flexible control across multiple dimensions, while still preserving high-relevance candidates.

The Future of Diversity-Aware Ranking

As we continue evolving our notification diversity-aware ranking system, a next step is to introduce more adaptive, dynamic demotion strategies. Instead of relying on static rules, we plan to make demotion strength responsive to notification volume and delivery timing. For example, as a user receives more notifications—especially of similar type or in rapid succession—the system progressively applies stronger penalties to new notification candidates, effectively mitigating overwhelming experiences caused by high notification volume or tightly spaced deliveries.

Longer term, we see an opportunity to bring large language models (LLMs) into the diversity pipeline. LLMs can help us go beyond surface-level rules by understanding semantic similarity between messages and rephrasing content in more varied, user-friendly ways. This would allow us to personalize notification experiences with richer language and improved relevance while maintaining diversity across topics, tone, and timing.

Source link

Events & Conferences

Simplifying book discovery with ML-powered visual autocomplete suggestions

Published

4 days ago

September 2, 2025

Mao Sheng Liu

Every day, millions of customers search for books in various formats (audiobooks, e-books, and physical books) across Amazon and Audible. Traditional keyword autocomplete suggestions, while helpful, usually require several steps before customers find their desired content. Audible took on the challenge of making book discovery more intuitive and personalized while reducing the number of steps to purchase.

We developed an instant visual autocomplete system that enhances the search experience across Amazon and Audible. As the user begins typing a query, our solution provides visual previews with book covers, enabling direct navigation to relevant landing pages instead of the search result page. It also delivers real-time personalized format recommendations and incorporates multiple searchable entities, such as book pages, author pages, and series pages.

1 of 2

Audible’s visual-autocomplete experience.

2 of 2

Amazon’s visual-autocomplete experience.

Our system needed to understand user intent from just a few keystrokes and determine the most relevant books to display, all while maintaining low latency for millions of queries. Using historical search data, we match keystrokes to products, transforming partial inputs into meaningful search suggestions. To ensure quality, we implemented confidence-based filtering mechanisms, which are particularly important for distinguishing between general queries like “mystery” and specific title searches. To reflect customers’ most recent interests, the system applies time-decay functions to long historical user interaction data.

Related content

Assessing the absolute utility of query results, rather than just their relative utility, improves learning-to-rank models.

To meet the unique requirements of each use case, we developed two distinct technical approaches. On Audible, we deployed a deep pairwise-learning-to-rank (DeepPLTR) model. The DeepPLTR model considers pairs of books and learns to assign a higher score to the one that better matches the customer query.

The DeepPLTR model’s architecture consists of three specialized towers. The left tower factors in contextual features and recent search patterns using a long-short-term-memory model, which processes data sequentially and considers its prior decisions when issuing a new term in the sequence. The middle tower handles keyword and item engagement history. The right tower factors in customer taste preferences and product descriptions to enable personalization. The model learns from paired examples, but at runtime, it relies on books’ absolute scores to assemble a ranked list.

Training architecture of the DeepPLTR model, which takes in paired examples (green and pink blocks). At runtime, the model scores only a single candidate at a time.

For Amazon, we implemented a two-stage modeling approach involving a probabilistic information-retrieval model to determine the book title that best matches each keyword and a second model that personalizes the book format (audiobooks, e-books, and physical books). This dual-strategy approach maintains low latency while still enabling personalization.

In practice, a customer who types “dungeon craw” in the search bar now sees a visual recommendation for the book Dungeon Crawler Carl, complete with book cover, reducing friction by bypassing a search results page and sending the customer directly to the product detail page. On Audible, the system also personalizes autocomplete results and enriches the discovery experience with relevant connections. These include links to the author’s complete works (Matt Dinniman’s author page) and, for titles that belong to a series, links to the full collection (such as the Dungeon Crawler Carl series).

Related content

Using reinforcement learning improves candidate selection and ranking for search, ad platforms, and recommender systems.

On Amazon, when the customer clicks on the title, the model personalizes the right book-format (audiobooks, e-books, physical books) recommendation and directs the customer to the right product detail page.

In both cases, after the customer has entered a certain number of keystrokes, the system employs a model to detect customer intent (e.g., book title intent for Amazon or author intent for Audible) and determine which visual widget should be displayed.

Audible and Amazon books’ visual autocomplete provides customers with more relevant content more rapidly than traditional autocomplete, and its direct navigation reduces the number of steps to find and access desired books — all while handling millions of queries at low latency.

This technology is not just about making book discovery easier; it is laying the foundation for future improvements in search personalization and visual discovery across Amazon’s ecosystem.

Acknowledgements: Jiun Kim, Sumit Khetan, Armen Stepanyan, Jack Xuan, Nathan Brothers, Eddie Chen, Vincent Lee, Soumy Ladha, Justine Luo, Yuchen Zeng, David Torres, Gali Deutsch, Chaitra Ramdas, Christopher Gomez, Sharmila Tamby, Melissa Ma, Cheng Luo, Jeffrey Jiang, Pavel Fedorov, Ronald Denaux, Aishwarya Vasanth, Azad Bajaj, Mary Heer, Adam Lowe, Jenny Wang, Cameron Cramer, Emmanuel Ankrah, Lydia Diaz, Suzette Islam, Fei Gu, Phil Weaver, Huan Xue, Kimmy Dai, Evangeline Yang, Chao Zhu, Anvy Tran, Jessica Wu, Xiaoxiong Huang, Jiushan Yang

Source link

Events & Conferences

Revolutionizing warehouse automation with scientific simulation

Published

2 weeks ago

August 26, 2025

Deniz Akyildiz

Modern warehouses rely on complex networks of sensors to enable safe and efficient operations. These sensors must detect everything from packages and containers to robots and vehicles, often in changing environments with varying lighting conditions. More important for Amazon, we need to be able to detect barcodes in an efficient way.

Related content

Generative AI supports the creation, at scale, of complex, realistic driving scenarios that can be directed to specific locations and environments.

The Amazon Robotics ID (ARID) team focuses on solving this problem. When we first started working on it, we faced a significant bottleneck: optimizing sensor placement required weeks or months of physical prototyping and real-world testing, severely limiting our ability to explore innovative solutions.

To transform this process, we developed Sensor Workbench (SWB), a sensor simulation platform built on NVIDIA’s Isaac Sim that combines parallel processing, physics-based sensor modeling, and high-fidelity 3-D environments. By providing virtual testing environments that mirror real-world conditions with unprecedented accuracy, SWB allows our teams to explore hundreds of configurations in the same amount of time it previously took to test just a few physical setups.

Camera and target selection/positioning

Sensor Workbench users can select different cameras and targets and position them in 3-D space to receive real-time feedback on barcode decodability.

Three key innovations enabled SWB: a specialized parallel-computing architecture that performs simulation tasks across the GPU; a custom CAD-to-OpenUSD (Universal Scene Description) pipeline; and the use of OpenUSD as the ground truth throughout the simulation process.