Large Language Model driven Policy Exploration for Recommender Systems

Recent advancements in Recommender Systems (RS) have incorporated Reinforcement Learning (RL), framing the recommendation as a Markov Decision Process (MDP). However, offline RL policies trained on static user data are vulnerable to distribution shift when deployed in dynamic online environments. Additionally, excessive focus on exploiting short-term relevant items can hinder exploration, leading to suboptimal recommendations and negatively impacting long-term user gains. Online RL-based RS also face challenges in production deployment, due to the risks of exposing users to untrained or unstable policies. Large Language Models (LLMs) offer a promising solution to mimic user objectives and preferences for pre-training policies offline to enhance the initial recommendations in online settings. Effectively managing distribution shift and balancing exploration are crucial for improving RL-based RS, especially when leveraging LLM-based pre-training. To address these challenges, we propose an Interaction-Augmented Learned Policy (iALP) that utilizes user preferences distilled from an LLM. Our approach involves prompting the LLM with user states to extract item preferences, learning rewards based on feedback, and updating the RL policy using an actor-critic framework. Furthermore, to deploy iALP in an online scenario, we introduce an adaptive variant, A-iALP, that implements a simple fine-tuning strategy (A-iALP$_{ft}$), and an adaptive approach (A-iALP$_{ap}$) designed to mitigate issues with compromised policies and limited exploration. Experiments across three simulated environments demonstrate that A-iALP introduces substantial performance improvements

Efficient and Effective Adaptation of Multimodal Foundation Models in Sequential Recommendation

Multimodal foundation models (MFMs) have revolutionized sequential recommender systems through advanced representation learning. While Parameter-efficient Fine-tuning (PEFT) is commonly used to adapt these models, studies often prioritize parameter efficiency, neglecting GPU memory and training speed. To address this, we introduced the IISAN framework, significantly enhancing efficiency. However, IISAN was limited to symmetrical MFMs and identical text and image encoders, preventing the use of state-of-the-art Large Language Models. To overcome this, we developed IISAN-Versa, a versatile plug-and-play architecture compatible with both symmetrical and asymmetrical MFMs. IISAN-Versa employs a Decoupled PEFT structure and utilizes both intra- and inter-modal adaptation. It effectively handles asymmetry through a simple yet effective combination of group layer-dropping and dimension transformation alignment. Our research demonstrates that IISAN-Versa effectively adapts large text encoders, and we further identify a scaling effect where larger encoders generally perform better. IISAN-Versa also demonstrates strong versatility in our defined multimodal scenarios, which include raw titles and captions generated from images and videos. Additionally, IISAN-Versa achieved state-of-the-art performance on the Microlens public benchmark. We will release our code and datasets to support future research.

Seeing Eye to AI: Human Alignment via Gaze-Based Response Rewards for Large Language Models

Advancements in Natural Language Processing (NLP), have led to the emergence of Large Language Models (LLMs) such as GPT, Llama, Claude, and Gemini, which excel across a range of tasks but require extensive fine-tuning to align their outputs with human expectations. A widely used method for achieving this alignment is Reinforcement Learning from Human Feedback (RLHF), which, despite its success, faces challenges in accurately modelling human preferences. In this paper, we introduce GazeReward, a novel framework that integrates implicit feedback -- and specifically eye-tracking (ET) data -- into the Reward Model (RM). In addition, we explore how ET-based features can provide insights into user preferences. Through ablation studies we test our framework with different integration methods, LLMs, and ET generator models, demonstrating that our approach significantly improves the accuracy of the RM on established human preference datasets. This work advances the ongoing discussion on optimizing AI alignment with human values, exploring the potential of cognitive data for shaping future NLP research.

PERSOMA: PERsonalized SOft ProMpt Adapter Architecture for Personalized Language Prompting

Understanding the nuances of a user's extensive interaction history is key to building accurate and personalized natural language systems that can adapt to evolving user preferences. To address this, we introduce PERSOMA, Personalized Soft Prompt Adapter architecture. Unlike previous personalized prompting methods for large language models, PERSOMA offers a novel approach to efficiently capture user history. It achieves this by resampling and compressing interactions as free form text into expressive soft prompt embeddings, building upon recent research utilizing embedding representations as input for LLMs. We rigorously validate our approach by evaluating various adapter architectures, first-stage sampling strategies, parameter-efficient tuning techniques like LoRA, and other personalization methods. Our results demonstrate PERSOMA's superior ability to handle large and complex user histories compared to existing embedding-based and text-prompt-based techniques.

IISAN: Efficiently Adapting Multimodal Representation for Sequential Recommendation with Decoupled PEFT

Multimodal foundation models are transformative in sequential recommender systems, leveraging powerful representation learning capabilities. While Parameter-efficient Fine-tuning (PEFT) is commonly used to adapt foundation models for recommendation tasks, most research prioritizes parameter efficiency, often overlooking critical factors like GPU memory efficiency and training speed. Addressing this gap, our paper introduces IISAN (Intra- and Inter-modal Side Adapted Network for Multimodal Representation), a simple plug-and-play architecture using a Decoupled PEFT structure and exploiting both intra- and inter-modal adaptation. IISAN matches the performance of full fine-tuning (FFT) and state-of-the-art PEFT. More importantly, it significantly reduces GPU memory usage - from 47GB to just 3GB for multimodal sequential recommendation tasks. Additionally, it accelerates training time per epoch from 443s to 22s compared to FFT. This is also a notable improvement over the Adapter and LoRA, which require 37-39 GB GPU memory and 350-380 seconds per epoch for training. Furthermore, we propose a new composite efficiency metric, TPME (Training-time, Parameter, and GPU Memory Efficiency) to alleviate the prevalent misconception that "parameter efficiency represents overall efficiency". TPME provides more comprehensive insights into practical efficiency comparisons between different methods. Besides, we give an accessible efficiency analysis of all PEFT and FFT approaches, which demonstrate the superiority of IISAN. We release our codes and other materials at https://github.com/GAIR-Lab/IISAN.

Reinforcement Learning-based Recommender Systems with Large Language Models for State Reward and Action Modeling

Reinforcement Learning (RL)-based recommender systems have demonstrated promising performance in meeting user expectations by learning to make accurate next-item recommendations from historical user-item interactions. However, existing offline RL-based sequential recommendation methods face the challenge of obtaining effective user feedback from the environment. Effectively modeling the user state and shaping an appropriate reward for recommendation remains a challenge. In this paper, we leverage language understanding capabilities and adapt large language models (LLMs) as an environment (LE) to enhance RL-based recommenders. The LE is learned from a subset of user-item interaction data, thus reducing the need for large training data, and can synthesise user feedback for offline data by: (i) acting as a state model that produces high quality states that enrich the user representation, and (ii) functioning as a reward model to accurately capture nuanced user preferences on actions. Moreover, the LE allows to generate positive actions that augment the limited offline training data. We propose a LE Augmentation (LEA) method to further improve recommendation performance by optimising jointly the supervised component and the RL policy, using the augmented actions and historical user signals. We use LEA, the state and reward models in conjunction with state-of-the-art RL recommenders and report experimental results on two publicly available datasets.

Latent User Intent Modeling for Sequential Recommenders

Sequential recommender models are essential components of modern industrial recommender systems. These models learn to predict the next items a user is likely to interact with based on his/her interaction history on the platform. Most sequential recommenders however lack a higher-level understanding of user intents, which often drive user behaviors online. Intent modeling is thus critical for understanding users and optimizing long-term user experience. We propose a probabilistic modeling approach and formulate user intent as latent variables, which are inferred based on user behavior signals using variational autoencoders (VAE). The recommendation policy is then adjusted accordingly given the inferred user intent. We demonstrate the effectiveness of the latent user intent modeling via offline analyses as well as live experiments on a large-scale industrial recommendation platform.

Rethinking Reinforcement Learning for Recommendation: A Prompt Perspective

Modern recommender systems aim to improve user experience. As reinforcement learning (RL) naturally fits this objective -- maximizing an user's reward per session -- it has become an emerging topic in recommender systems. Developing RL-based recommendation methods, however, is not trivial due to the \emph{offline training challenge}. Specifically, the keystone of traditional RL is to train an agent with large amounts of online exploration making lots of `errors' in the process. In the recommendation setting, though, we cannot afford the price of making `errors' online. As a result, the agent needs to be trained through offline historical implicit feedback, collected under different recommendation policies; traditional RL algorithms may lead to sub-optimal policies under these offline training settings. Here we propose a new learning paradigm -- namely Prompt-Based Reinforcement Learning (PRL) -- for the offline training of RL-based recommendation agents. While traditional RL algorithms attempt to map state-action input pairs to their expected rewards (e.g., Q-values), PRL directly infers actions (i.e., recommended items) from state-reward inputs. In short, the agents are trained to predict a recommended item given the prior interactions and an observed reward value -- with simple supervised learning. At deployment time, this historical (training) data acts as a knowledge base, while the state-reward pairs are used as a prompt. The agents are thus used to answer the question: \emph{ Which item should be recommended given the prior interactions \& the prompted reward value}? We implement PRL with four notable recommendation models and conduct experiments on two real-world e-commerce datasets. Experimental results demonstrate the superior performance of our proposed methods.

Enhancing Top-N Item Recommendations by Peer Collaboration

Deep neural networks (DNN) have achieved great success in the recommender systems (RS) domain. However, to achieve remarkable performance, DNN-based recommender models often require numerous parameters, which inevitably bring redundant neurons and weights, a phenomenon referred to as over-parameterization. In this paper, we plan to exploit such redundancy phenomena to improve the performance of RS. Specifically, we propose PCRec, a top-N item \underline{rec}ommendation framework that leverages collaborative training of two DNN-based recommender models with the same network structure, termed \underline{p}eer \underline{c}ollaboration. PCRec can reactivate and strengthen the unimportant (redundant) weights during training, which achieves higher prediction accuracy but maintains its original inference efficiency. To realize this, we first introduce two criteria to identify the importance of weights of a given recommender model. Then, we rejuvenate the unimportant weights by transplanting outside information (i.e., weights) from its peer network. After such an operation and retraining, the original recommender model is endowed with more representation capacity by possessing more functional model parameters. To show its generality, we instantiate PCRec by using three well-known recommender models. We conduct extensive experiments on three real-world datasets, and show that PCRec yields significantly better recommendations than its counterpart with the same model (parameter) size.

Supervised Advantage Actor-Critic for Recommender Systems

Casting session-based or sequential recommendation as reinforcement learning (RL) through reward signals is a promising research direction towards recommender systems (RS) that maximize cumulative profits. However, the direct use of RL algorithms in the RS setting is impractical due to challenges like off-policy training, huge action spaces and lack of sufficient reward signals. Recent RL approaches for RS attempt to tackle these challenges by combining RL and (self-)supervised sequential learning, but still suffer from certain limitations. For example, the estimation of Q-values tends to be biased toward positive values due to the lack of negative reward signals. Moreover, the Q-values also depend heavily on the specific timestamp of a sequence. To address the above problems, we propose negative sampling strategy for training the RL component and combine it with supervised sequential learning. We call this method Supervised Negative Q-learning (SNQN). Based on sampled (negative) actions (items), we can calculate the "advantage" of a positive action over the average case, which can be further utilized as a normalized weight for learning the supervised sequential part. This leads to another learning framework: Supervised Advantage Actor-Critic (SA2C). We instantiate SNQN and SA2C with four state-of-the-art sequential recommendation models and conduct experiments on two real-world datasets. Experimental results show that the proposed approaches achieve significantly better performance than state-of-the-art supervised methods and existing self-supervised RL methods . Code will be open-sourced.