arXiv subjects

Yesu Feng

Publications and source records attributed to Yesu Feng.

9 recordsLinked to original sources

arXiv2026

From Logs to Language: Learning Optimal Verbalization for LLM-Based Recommendation at Industry Scale

Large language models (LLMs) are promising backbones for generative recommender systems, yet a key challenge remains underexplored: verbalization, i.e., converting structured user interaction logs into effective natural language inputs. Existing methods rely on rigid templates that simply concatenate fields, yielding suboptimal representations for recommendation. We propose a data-centric framework that learns verbalization for LLM-based recommendation. Using reinforcement learning, a verbalization agent transforms raw interaction histories into optimized textual contexts, with recommendation accuracy as the training signal. This agent learns to filter noise, incorporate relevant metadata, and reorganize information to improve downstream predictions. Experiments on a large-scale industrial streaming dataset from Netflix show that learned verbalization delivers up to 93% relative improvement in discovery item recommendation accuracy over template-based baselines. Further analysis reveals emergent strategies such as user interest summarization, noise removal, and syntax normalization, offering insights into effective context construction for LLM-based recommender systems.

cs.AI↗

arXiv2026

Netflix Artwork Personalization via LLM Post-training

Large language models (LLMs) have demonstrated success in various applications of user recommendation and personalization across e-commerce and entertainment. On many entertainment platforms such as Netflix, users typically interact with a wide range of titles, each represented by an artwork. Since users have diverse preferences, an artwork that appeals to one type of user may not resonate with another with different preferences. Given this user heterogeneity, our work explores the novel problem of personalized artwork recommendations according to diverse user preferences. Similar to the multi-dimensional nature of users' tastes, titles contain different themes and tones that may appeal to different viewers. For example, the same title might feature both heartfelt family drama and intense action scenes. Users who prefer romantic content may like the artwork emphasizing emotional warmth between the characters, while those who prefer action thrillers may find high-intensity action scenes more intriguing. Rather than a one-size-fits-all approach, we conduct post-training of pre-trained LLMs to make personalized artwork recommendations, selecting the most preferred visual representation of a title for each user and thereby improving user satisfaction and engagement. Our experimental results with Llama 3.1 8B models (trained on a dataset of 110K data points and evaluated on 5K held-out user-title pairs) show that the post-trained LLMs achieve 3-5\% improvements over the Netflix production model, suggesting a promising direction for granular personalized recommendations using LLMs.

cs.IR↗

arXiv2025

Collaborative Retrieval for Large Language Model-based Conversational Recommender Systems

Conversational recommender systems (CRS) aim to provide personalized recommendations via interactive dialogues with users. While large language models (LLMs) enhance CRS with their superior understanding of context-aware user preferences, they typically struggle to leverage behavioral data, which have proven to be important for classical collaborative filtering (CF)-based approaches. For this reason, we propose CRAG, Collaborative Retrieval Augmented Generation for LLM-based CRS. To the best of our knowledge, CRAG is the first approach that combines state-of-the-art LLMs with CF for conversational recommendations. Our experiments on two publicly available movie conversational recommendation datasets, i.e., a refined Reddit dataset (which we name Reddit-v2) as well as the Redial dataset, demonstrate the superior item coverage and recommendation performance of CRAG, compared to several CRS baselines. Moreover, we observe that the improvements are mainly due to better recommendation accuracy on recently released movies. The code and data are available at https://github.com/yaochenzhu/CRAG.

cs.IR↗

arXiv2024

Sliding Window Training -- Utilizing Historical Recommender Systems Data for Foundation Models

Long-lived recommender systems (RecSys) often encounter lengthy user-item interaction histories that span many years. To effectively learn long term user preferences, Large RecSys foundation models (FM) need to encode this information in pretraining. Usually, this is done by either generating a long enough sequence length to take all history sequences as input at the cost of large model input dimension or by dropping some parts of the user history to accommodate model size and latency requirements on the production serving side. In this paper, we introduce a sliding window training technique to incorporate long user history sequences during training time without increasing the model input dimension. We show the quantitative & qualitative improvements this technique brings to the RecSys FM in learning user long term preferences. We additionally show that the average quality of items in the catalog learnt in pretraining also improves.

cs.IR↗

arXiv2024

IntentRec: Predicting User Session Intent with Hierarchical Multi-Task Learning

Recommender systems have played a critical role in diverse digital services such as e-commerce, streaming media, social networks, etc. If we know what a user's intent is in a given session (e.g. do they want to watch short videos or a movie or play games; are they shopping for a camping trip), it becomes easier to provide high-quality recommendations. In this paper, we introduce IntentRec, a novel recommendation framework based on hierarchical multi-task neural network architecture that tries to estimate a user's latent intent using their short- and long-term implicit signals as proxies and uses the intent prediction to predict the next item user is likely to engage with. By directly leveraging the intent prediction, we can offer accurate and personalized recommendations to users. Our comprehensive experiments on Netflix user engagement data show that IntentRec outperforms the state-of-the-art next-item and next-intent predictors. We also share several findings and downstream applications of IntentRec.

cs.IR↗

arXiv2023

Large Language Models as Zero-Shot Conversational Recommenders

In this paper, we present empirical studies on conversational recommendation tasks using representative large language models in a zero-shot setting with three primary contributions. (1) Data: To gain insights into model behavior in "in-the-wild" conversational recommendation scenarios, we construct a new dataset of recommendation-related conversations by scraping a popular discussion website. This is the largest public real-world conversational recommendation dataset to date. (2) Evaluation: On the new dataset and two existing conversational recommendation datasets, we observe that even without fine-tuning, large language models can outperform existing fine-tuned conversational recommendation models. (3) Analysis: We propose various probing tasks to investigate the mechanisms behind the remarkable performance of large language models in conversational recommendation. We analyze both the large language models' behaviors and the characteristics of the datasets, providing a holistic understanding of the models' effectiveness, limitations and suggesting directions for the design of future conversational recommenders

cs.IR↗

arXiv2017

Prostate Cancer Diagnosis using Deep Learning with 3D Multiparametric MRI

A novel deep learning architecture (XmasNet) based on convolutional neural networks was developed for the classification of prostate cancer lesions, using the 3D multiparametric MRI data provided by the PROSTATEx challenge. End-to-end training was performed for XmasNet, with data augmentation done through 3D rotation and slicing, in order to incorporate the 3D information of the lesion. XmasNet outperformed traditional machine learning models based on engineered features, for both train and test data. For the test data, XmasNet outperformed 69 methods from 33 participating groups and achieved the second highest AUC (0.84) in the PROSTATEx challenge. This study shows the great potential of deep learning for cancer imaging.

cs.CV↗

arXiv2013

Measuring Long-Lived ^{13}C-Singlet State Lifetimes at Natural Abundance

Long-lived singlet states hold the potential to drastically extend the lifetime of hyperpolarization in molecular tracers for in-vivo magnetic resonance imaging (MRI). Such long lived hyperpolarization can be used for elucidation of fundamental metabolic pathways, early diagnosis, and optimization of clinical tests for new medication. All previous measurements of 13C singlet state lifetimes rely on costly and time consuming syntheses of 13C labeled compounds. Here we show that it is possible to determine 13C singlet state lifetimes by detecting the naturally abundant doubly-labeled species. This approach allows for rapid and low cost screening of potential molecular biomarkers bearing long-lived singlet states.

physics.chem-ph↗

arXiv2013

Spin lock composite and shaped pulses for efficient and robust pumping of dark states in magnetic resonance

Long-lived (symmetry protected) hyperpolarized spin states offer important new opportunities (for example, in clinical MR imaging), but existing methods for producing these states are limited by either excess energy dissipation or high sensitivity to inhomogeneities. We extend recent work on continuous-wave irradiation of nearly-equivalent spins (spin-lock induced crossing) by designing composite pulse and adiabatic shaped-pulse excitations which overcome the limitations. These composite and adiabatic pulses differ drastically from the traditional solutions in two-level systems. We also show this works in chemically equivalent spin pairs, which has the advantage of allowing for polarization transfer from and to remote spins. The approach is broadly applicable to systems where varying excitation strength induces an avoided crossing to a dark state, and thus to many other spectroscopic regimes.

physics.chem-ph↗