Heng-Tze Cheng
Heng-Tze Cheng is a Technical Lead Manager and Senior Staff Software Engineer on the Google Brain team, part of Google Research & AI. Heng-Tze currently leads a research team focusing on Neural Sequence Modeling research for Task-oriented Dialogues, Personalized Semantic Search, and Recommender Systems productionized across Google, such as Google Duplex Assistant, YouTube, and more. Heng-Tze also founded and led the Wide & Deep Learning project in TensorFlow, and has worked on large-scale machine learning platforms that are widely used for retrieval, ranking, and recommender systems.
Prior to joining Google in 2014, Heng-Tze received his Ph.D. from Carnegie Mellon University in 2013 and B.S. from National Taiwan University in 2008. His research interests range across machine learning, information retrieval, user behavior modeling, and human activity recognition.
Authored Publications
Sort By
HyperPrompt: Prompt-based Task-Conditioning of Transformers
Cosmo Du
Steven Zheng
Vamsi Aribandi
Yi Tay
Yun He
Zhao Chen
Zhe Zhao
ICML (2022)
Preview abstract
Prompt-tuning is becoming a new paradigm for finetuning pre-trained language models in a parameter-efficient way. Here, we explore the use of HyperNetworks to generate prompts. We propose a novel architecture of HyperPrompt: prompt-based task-conditioned parameterization of self-attention in Transformers. We show that HyperPrompt is very competitive against strong multi-task learning baselines with only 1% of additional task-conditioning parameters. The prompts are end-to-end learnable via generation by a HyperNetwork. The additional parameters scale sub-linearly with the number of downstream tasks, which makes it very parameter efficient for multi-task learning. Hyper-Prompt allows the network to learn task-specific feature maps where the prompts serve as task global memories. Information sharing is enabled among tasks through the HyperNetwork to alleviate task conflicts during co-training. Through extensive empirical experiments, we demonstrate that HyperPrompt can achieve superior performances over strong T5 multi-task learning base-lines and parameter-efficient adapter variants including Prompt-Tuning on Natural Language Understanding benchmarks of GLUE and Super-GLUE across all the model sizes explored.
View details
LaMDA: Language Models for Dialog Applications
Aaron Daniel Cohen
Alena Butryna
Alicia Jin
Apoorv Kulshreshtha
Ben Zevenbergen
Chung-ching Chang
Cosmo Du
Daniel De Freitas Adiwardana
Dehao Chen
Dmitry (Dima) Lepikhin
Erin Hoffman-John
Igor Krivokon
James Qin
Jamie Hall
Joe Fenton
Johnny Soraker
Kathy Meier-Hellstern
Maarten Paul Bosma
Marc Joseph Pickett
Marcelo Amorim Menegali
Marian Croak
Maxim Krikun
Noam Shazeer
Rachel Bernstein
Ravi Rajakumar
Ray Kurzweil
Romal Thoppilan
Steven Zheng
Taylor Bos
Toju Duke
Tulsee Doshi
Vincent Y. Zhao
Will Rusch
Yuanzhong Xu
arXiv (2022)
Preview abstract
We present LaMDA: Language Models for Dialog Applications. LaMDA is a family of Transformer-based neural language models specialized for dialog, which have up to 137B parameters and arepre-trained on 1.56T words of public dialog data and web text. While model scaling alone canimprove quality, it shows less improvements on safety and factual grounding. We demonstrate thatfine-tuning with annotated data and enabling the model to consult external knowledge sources canlead to significant improvements towards the two key challenges of safety and factual grounding.The first challenge, safety, involves ensuring that the model’s responses are consistent with a set ofhuman values, such as preventing harmful suggestions and unfair bias. We quantify safety using ametric based on an illustrative set of values, and we find that filtering candidate responses using aLaMDA classifier fine-tuned with a small amount of crowdworker-annotated data offers a promisingapproach to improving model safety. The second challenge, factual grounding, involves enabling themodel to consult external knowledge sources, such as an information retrieval system, a languagetranslator, and a calculator. We quantify factuality using a groundedness metric, and we find that ourapproach enables the model to generate responses grounded in known sources, rather than responsesthat merely sound plausible. Finally, we explore the use of LaMDA in the domains of education andcontent recommendations, and analyze their helpfulness and role consistency.
View details
Mondegreen: A Post-Processing Solution to Speech Recognition Error Correction for Voice Search Queries
Ajit Apte
Ambarish Jash
Amol H Wankhede
Ankit Kumar
Ayooluwakunmi Jeje
Dima Kuzmin
Ellie Ka In Chio
Harry Fung
Jon Effrat
Nitin Jindal
Pei Cao
Senqiang Zhou
Sukhdeep S. Sodhi
Tameen Khan
Tarush Bali
KDD (2021)
Preview abstract
As more and more online search queries come from voice, automatic speech recognition becomes a key component to deliver relevant search results. Errors introduced by automatic speech recognition (ASR) lead to irrelevant search results returned to the user, thus causing user dissatisfaction. In this paper, we introduce an approach, Mondegreen, to correct voice queries in text space without depending on audio signals, which may not always be available due to system constraints or privacy or bandwidth (for example, some ASR systems run on-device) considerations. We focus on voice queries transcribed via several proprietary commercial ASR systems. These queries come from users making internet, or online service search queries. We first present an analysis showing how different the language distribution coming from user voice queries is from that in traditional text corpora used to train off-the-shelf ASR systems. We then demonstrate that Mondegreen can achieve significant improvements in increased user interaction by correcting user voice queries in one of the largest search systems in Google. Finally, we see Mondegreen as complementing existing highly-optimized production ASR systems, which may not be frequently retrained and thus lag behind due to vocabulary drifts.
View details
Zero-Shot Transfer Learning for Query-Item Cold Start in Search Retrieval and Recommendations
Ankit Kumar
Cosmo Du
Dima Kuzmin
Ellie Chio
John Roberts Anderson
Li Zhang
Nitin Jindal
Pei Cao
Ritesh Agarwal
Tao Wu
Wen Li
CIKM (2020)
Preview abstract
Most search retrieval and recommender systems predict top-K items given a query by learning directly from a large training set of (query, item) pairs, where a query can include natural language (NL), user, and context features. These approaches fall into the traditional supervised learning framework where the algorithm trains on labeled data from the target task. In this paper, we propose a new zero-shot transfer learning framework, which first learns representations of items and their NL features by predicting (item, item) correlation graphs as an auxiliary task, followed by transferring learned representations to solve the target task (query-to-item prediction), without having seen any (query, item) pairs in training. The advantages of applying this new framework include: (1) Cold-starting search and recommenders without abundant query-item data; (2) Generalizing to previously unseen or rare (query, item) pairs and alleviating the "rich get richer" problem; (3) Transferring knowledge of (item, item) correlation from domains outside of search. We show that the framework is effective on a large-scale search and recommender system.
View details
Reinforcement Learning for Slate-based Recommender Systems: A Tractable Decomposition and Practical Methodology
Vihan Jain
Jing Wang
Sanmit Narvekar
Ritesh Agarwal
Rui Wu
Morgane Lustman
Vince Gatto
Paul Covington
Jim McFadden
arXiv (2019)
Preview abstract
Most practical recommender systems focus on estimating immediate user engagement without considering the long-term effects of recommendations on user behavior. Reinforcement learning (RL) methods offer the potential to optimize recommendations for long-term user engagement. However, since users are often presented with slates of multiple items---which may have interacting effects on user choice---methods are required to deal with the combinatorics of the RL action space. In this work, we address the challenge of making slate-based recommendations to optimize long-term value using RL. Our contributions are three-fold. (i) We develop SlateQ, a decomposition of value-based temporal-difference and Q-learning that renders RL tractable with slates. Under mild assumptions on user-choice behavior, we show that the long-term value (LTV) of a slate can be decomposed into a tractable function of its component item-wise LTVs. (ii) We outline a methodology that leverages existing myopic learning-based recommenders to quickly develop a recommender that handles LTV. (iii) We demonstrate our methods in simulation, and validate the scalability of decomposed TD-learning using SlateQ in live experiments on YouTube.
View details
SlateQ: A Tractable Decomposition for Reinforcement Learning with Recommendation Sets
Vihan Jain
Jing Wang
Sanmit Narvekar
Ritesh Agarwal
Rui Wu
Proceedings of the Twenty-eighth International Joint Conference on Artificial Intelligence (IJCAI-19), Macau, China (2019), pp. 2592-2599
Preview abstract
Reinforcement learning (RL) methods for recommender systems optimize recommendations for long-term user engagement. However, since users are often presented with slates of multiple items---which may have interacting effects on user choice---methods are required to deal with the combinatorics of the RL action space. We develop SlateQ, a decomposition of value-based temporal-difference and Q-learning that renders RL tractable with slates. Under mild assumptions on user choice behavior, we show that the long-term value (LTV) of a slate can be decomposed into a tractable function of its component item-wise LTVs. We demonstrate our methods in simulation, and validate the scalability and effectiveness of decomposed TD-learning on YouTube.
View details
TensorFlow Estimators: Managing Simplicity vs. Flexibility in High-Level Machine Learning Frameworks
Cassandra Xia
Clemens Mewald
George Roumpos
Illia Polosukhin
Jamie Alexander Smith
Jianwei Xie
Lichan Hong
Mustafa Ispir
Philip Daniel Tucker
Yuan Tang
Proceedings of the 23th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Halifax, Canada (2017)
Preview abstract
We present a framework for specifying, training, evaluating, and deploying machine learning models. Our focus is to simplify writing cutting edge machine learning models in a way that enables bringing those models into production. Recognizing the fast evolution of the field of deep learning, we make no attempt to capture the design space of all possible model architectures in a DSL or similar configuration. We allow users to write code to define their models, but provide abstractions that guide developers to write models in ways conducive to productionization, as well as providing a unifying Estimator interface, a unified interface making it possible to write downstream infrastructure (distributed training, hyperparameter tuning, …) independent of the model implementation.
We balance the competing demands for flexibility and simplicity by offering APIs at different levels of abstraction, making common model architectures available “out of the box”, while providing a library of utilities designed to speed up experimentation with model architectures. To make out of the box models flexible and usable across a wide range of problems, these canned Estimators are parameterized not only over traditional hyperparameters, but also using feature columns, a declarative specification describing how to interpret input data.
We discuss our experience in using this framework in research and production environments, and show the impact on code health, maintainability, and development speed.
View details
TFX: A TensorFlow-Based Production-Scale Machine Learning Platform
Akshay Naresh Modi
Chiu Yuen Koo
Chuan Yu Foo
Clemens Mewald
Denis M. Baylor
Jarek Wilkiewicz
Levent Koc
Lukasz Lew
Martin A. Zinkevich
Mustafa Ispir
Neoklis Polyzotis
Steven Whang
Sudip Roy
Sukriti Ramesh
Vihan Jain
Xin Zhang
KDD 2017
Preview abstract
Creating and maintaining a platform for reliably producing and deploying machine learning models requires careful orchestration of many components—a learner for generating models based on training data, modules for analyzing and validating both data as well as models, and finally infrastructure for serving models in production. This becomes particularly challenging when data changes over time and fresh models need to be produced continuously. Unfortunately, such orchestration is often done ad hoc using glue code and custom scripts developed by individual teams for specific use cases, leading to duplicated effort and fragile systems with high technical debt.
We present TensorFlow Extended (TFX), a TensorFlow-based general-purpose machine learning platform implemented at Google. By integrating the aforementioned components into one platform, we were able to standardize the components, simplify the platform configuration, and reduce the time to production from the order of months to weeks, while providing platform stability that minimizes disruptions.
We present the case study of one deployment of TFX in the Google Play app store, where the machine learning models are refreshed continuously as new data arrive. Deploying TFX led to reduced custom code, faster experiment cycles, and a 2% increase in app installs resulting from improved data and model analysis.
View details
Wide & Deep Learning for Recommender Systems
Levent Koc
Tal Shaked
Glen Anderson
Wei Chai
Mustafa Ispir
Rohan Anil
Lichan Hong
Vihan Jain
Xiaobing Liu
Hemal Shah
arXiv:1606.07792 (2016)
Preview abstract
Generalized linear models with nonlinear feature transformations are widely used for large-scale regression and classification problems with sparse inputs. Memorization of feature interactions through a wide set of cross-product feature transformations are effective and interpretable, while generalization requires more feature engineering effort. With less feature engineering, deep neural networks can generalize better to unseen feature combinations through low-dimensional dense embeddings learned for the sparse features. However, deep neural networks with embeddings can over-generalize and recommend less relevant items when the user-item interactions are sparse and high-rank. In this paper, we present Wide & Deep learning---jointly trained wide linear models and deep neural networks---to combine the benefits of memorization and generalization for recommender systems. We productionized and evaluated the system on a commercial mobile app store with over one billion active users and over one million apps. Online experiment results show that Wide & Deep significantly increased app acquisitions compared with wide-only and deep-only models.
View details
