MARFT: Multi-Agent Reinforcement Fine-Tuning

LLM-based Multi-Agent Systems have demonstrated remarkable capabilities in addressing complex, agentic tasks requiring multifaceted reasoning and collaboration, from generating high-quality presentation slides to conducting sophisticated scientific research. Meanwhile, RL has been widely recognized for its effectiveness in enhancing agent intelligence, but limited research has investigated the fine-tuning of LaMAS using foundational RL techniques. Moreover, the direct application of MARL methodologies to LaMAS introduces significant challenges, stemming from the unique characteristics and mechanisms inherent to LaMAS. To address these challenges, this article presents a comprehensive study of LLM-based MARL and proposes a novel paradigm termed Multi-Agent Reinforcement Fine-Tuning (MARFT). We introduce a universal algorithmic framework tailored for LaMAS, outlining the conceptual foundations, key distinctions, and practical implementation strategies. We begin by reviewing the evolution from RL to Reinforcement Fine-Tuning, setting the stage for a parallel analysis in the multi-agent domain. In the context of LaMAS, we elucidate critical differences between MARL and MARFT. These differences motivate a transition toward a novel, LaMAS-oriented formulation of RFT. Central to this work is the presentation of a robust and scalable MARFT framework. We detail the core algorithm and provide a complete, open-source implementation to facilitate adoption and further research. The latter sections of the paper explore real-world application perspectives and opening challenges in MARFT. By bridging theoretical underpinnings with practical methodologies, this work aims to serve as a roadmap for researchers seeking to advance MARFT toward resilient and adaptive solutions in agentic systems. Our implementation of the proposed framework is publicly available at: https://github.com/jwliao-ai/MARFT.

A Survey of AI Agent Protocols

The rapid development of large language models (LLMs) has led to the widespread deployment of LLM agents across diverse industries, including customer service, content generation, data analysis, and even healthcare. However, as more LLM agents are deployed, a major issue has emerged: there is no standard way for these agents to communicate with external tools or data sources. This lack of standardized protocols makes it difficult for agents to work together or scale effectively, and it limits their ability to tackle complex, real-world tasks. A unified communication protocol for LLM agents could change this. It would allow agents and tools to interact more smoothly, encourage collaboration, and triggering the formation of collective intelligence. In this paper, we provide a systematic overview of existing communication protocols for LLM agents. We classify them into four main categories and make an analysis to help users and developers select the most suitable protocols for specific applications. Additionally, we conduct a comparative performance analysis of these protocols across key dimensions such as security, scalability, and latency. Finally, we explore future challenges, such as how protocols can adapt and survive in fast-evolving environments, and what qualities future protocols might need to support the next generation of LLM agent ecosystems. We expect this work to serve as a practical reference for both researchers and engineers seeking to design, evaluate, or integrate robust communication infrastructures for intelligent agents.

Information-Theoretic Reward Decomposition for Generalizable RLHF

A generalizable reward model is crucial in Reinforcement Learning from Human Feedback (RLHF) as it enables correctly evaluating unseen prompt-response pairs. However, existing reward models lack this ability, as they are typically trained by increasing the reward gap between chosen and rejected responses, while overlooking the prompts that the responses are conditioned on. Consequently, when the trained reward model is evaluated on prompt-response pairs that lie outside the data distribution, neglecting the effect of prompts may result in poor generalization of the reward model. To address this issue, we decompose the reward value into two independent components: prompt-free reward and prompt-related reward. Prompt-free reward represents the evaluation that is determined only by responses, while the prompt-related reward reflects the reward that derives from both the prompt and the response. We extract these two components from an information-theoretic perspective, which requires no extra models. Subsequently, we propose a new reward learning algorithm by prioritizing data samples based on their prompt-free reward values. Through toy examples, we demonstrate that the extracted prompt-free and prompt-related rewards effectively characterize two parts of the reward model. Further, standard evaluations show that our method improves both the alignment performance and the generalization capability of the reward model.

AdvKT: An Adversarial Multi-Step Training Framework for Knowledge Tracing

Knowledge Tracing (KT) monitors students' knowledge states and simulates their responses to question sequences. Existing KT models typically follow a single-step training paradigm, which leads to discrepancies with the multi-step inference process required in real-world simulations, resulting in significant error accumulation. This accumulation of error, coupled with the issue of data sparsity, can substantially degrade the performance of recommendation models in the intelligent tutoring systems. To address these challenges, we propose a novel Adversarial Multi-Step Training Framework for Knowledge Tracing (AdvKT), which, for the first time, focuses on the multi-step KT task. More specifically, AdvKT leverages adversarial learning paradigm involving a generator and a discriminator. The generator mimics high-reward responses, effectively reducing error accumulation across multiple steps, while the discriminator provides feedback to generate synthetic data. Additionally, we design specialized data augmentation techniques to enrich the training data with realistic variations, ensuring that the model generalizes well even in scenarios with sparse data. Experiments conducted on four real-world datasets demonstrate the superiority of AdvKT over existing KT models, showcasing its ability to address both error accumulation and data sparsity issues effectively.

AgentNet: Decentralized Evolutionary Coordination for LLM-based Multi-Agent Systems

The rapid advancement of Large Language Models (LLMs) has catalyzed the development of multi-agent systems, where multiple LLM-based agents collaborate to solve complex tasks. However, existing systems predominantly rely on centralized coordination, which introduces scalability bottlenecks, limits adaptability, and creates single points of failure. Additionally, concerns over privacy and proprietary knowledge sharing hinder cross-organizational collaboration, leading to siloed expertise. To address these challenges, we propose AgentNet, a decentralized, Retrieval-Augmented Generation (RAG)-based framework that enables LLM-based agents to autonomously evolve their capabilities and collaborate efficiently in a Directed Acyclic Graph (DAG)-structured network. Unlike traditional multi-agent systems that depend on static role assignments or centralized control, AgentNet allows agents to specialize dynamically, adjust their connectivity, and route tasks without relying on predefined workflows. AgentNet's core design is built upon several key innovations: (1) Fully Decentralized Paradigm: Removing the central orchestrator, allowing agents to coordinate and specialize autonomously, fostering fault tolerance and emergent collective intelligence. (2) Dynamically Evolving Graph Topology: Real-time adaptation of agent connections based on task demands, ensuring scalability and resilience.(3) Adaptive Learning for Expertise Refinement: A retrieval-based memory system that enables agents to continuously update and refine their specialized skills. By eliminating centralized control, AgentNet enhances fault tolerance, promotes scalable specialization, and enables privacy-preserving collaboration across organizations. Through decentralized coordination and minimal data exchange, agents can leverage diverse knowledge sources while safeguarding sensitive information.

Sell It Before You Make It: Revolutionizing E-Commerce with Personalized AI-Generated Items

E-commerce has revolutionized retail, yet its traditional workflows remain inefficient, with significant time and resource costs tied to product design and manufacturing inventory. This paper introduces a novel system deployed at Alibaba that leverages AI-generated items (AIGI) to address these challenges with personalized text-to-image generation for e-commercial product design. AIGI enables an innovative business mode called "sell it before you make it", where merchants can design fashion items and generate photorealistic images with digital models based on textual descriptions. Only when the items have received a certain number of orders, do the merchants start to produce them, which largely reduces reliance on physical prototypes and thus accelerates time to market. For such a promising application, we identify the underlying key scientific challenge, i.e., capturing the users' group-level personalized preferences towards multiple generated candidate images. To this end, we propose a Personalized Group-Level Preference Alignment Framework for Diffusion Models (i.e., PerFusion). We first design PerFusion Reward Model for user preference estimation with a feature-crossing-based personalized plug-in. Then we develop PerFusion with a personalized adaptive network to model diverse preferences across users, and meanwhile derive the group-level preference optimization objective to capture the comparative behaviors among multiple candidates. Both offline and online experiments demonstrate the effectiveness of our proposed algorithm. The AI-generated items have achieved over 13% relative improvements for both click-through rate and conversion rate compared to their human-designed counterparts, validating the revolutionary potential of AI-generated items for e-commercial platforms.

ReMA: Learning to Meta-think for LLMs with Multi-Agent Reinforcement Learning

Recent research on Reasoning of Large Language Models (LLMs) has sought to further enhance their performance by integrating meta-thinking -- enabling models to monitor, evaluate, and control their reasoning processes for more adaptive and effective problem-solving. However, current single-agent work lacks a specialized design for acquiring meta-thinking, resulting in low efficacy. To address this challenge, we introduce Reinforced Meta-thinking Agents (ReMA), a novel framework that leverages Multi-Agent Reinforcement Learning (MARL) to elicit meta-thinking behaviors, encouraging LLMs to think about thinking. ReMA decouples the reasoning process into two hierarchical agents: a high-level meta-thinking agent responsible for generating strategic oversight and plans, and a low-level reasoning agent for detailed executions. Through iterative reinforcement learning with aligned objectives, these agents explore and learn collaboration, leading to improved generalization and robustness. Experimental results demonstrate that ReMA outperforms single-agent RL baselines on complex reasoning tasks, including competitive-level mathematical benchmarks and LLM-as-a-Judge benchmarks. Comprehensive ablation studies further illustrate the evolving dynamics of each distinct agent, providing valuable insights into how the meta-thinking reasoning process enhances the reasoning capabilities of LLMs.

Adding Alignment Control to Language Models

Post-training alignment has increasingly become a crucial factor in enhancing the usability of language models (LMs). However, the strength of alignment varies depending on individual preferences. This paper proposes a method to incorporate alignment control into a single model, referred to as CLM. This approach adds one identity layer preceding the initial layers and performs preference learning only on this layer to map unaligned input token embeddings into the aligned space. Experimental results demonstrate that this efficient fine-tuning method performs comparable to full fine-tuning. During inference, the input embeddings are processed through the aligned and unaligned layers, which are then merged through the interpolation coefficient. By controlling this parameter, the alignment exhibits a clear interpolation and extrapolation phenomenon.

PALo: Learning Posture-Aware Locomotion for Quadruped Robots

With the rapid development of embodied intelligence, locomotion control of quadruped robots on complex terrains has become a research hotspot. Unlike traditional locomotion control approaches focusing solely on velocity tracking, we pursue to balance the agility and robustness of quadruped robots on diverse and complex terrains. To this end, we propose an end-to-end deep reinforcement learning framework for posture-aware locomotion named PALo, which manages to handle simultaneous linear and angular velocity tracking and real-time adjustments of body height, pitch, and roll angles. In PALo, the locomotion control problem is formulated as a partially observable Markov decision process, and an asymmetric actor-critic architecture is adopted to overcome the sim-to-real challenge. Further, by incorporating customized training curricula, PALo achieves agile posture-aware locomotion control in simulated environments and successfully transfers to real-world settings without fine-tuning, allowing real-time control of the quadruped robot's locomotion and body posture across challenging terrains. Through in-depth experimental analysis, we identify the key components of PALo that contribute to its performance, further validating the effectiveness of the proposed method. The results of this study provide new possibilities for the low-level locomotion control of quadruped robots in higher dimensional command spaces and lay the foundation for future research on upper-level modules for embodied intelligence.

Humanoid Whole-Body Locomotion on Narrow Terrain via Dynamic Balance and Reinforcement Learning

Humans possess delicate dynamic balance mechanisms that enable them to maintain stability across diverse terrains and under extreme conditions. However, despite significant advances recently, existing locomotion algorithms for humanoid robots are still struggle to traverse extreme environments, especially in cases that lack external perception (e.g., vision or LiDAR). This is because current methods often rely on gait-based or perception-condition rewards, lacking effective mechanisms to handle unobservable obstacles and sudden balance loss. To address this challenge, we propose a novel whole-body locomotion algorithm based on dynamic balance and Reinforcement Learning (RL) that enables humanoid robots to traverse extreme terrains, particularly narrow pathways and unexpected obstacles, using only proprioception. Specifically, we introduce a dynamic balance mechanism by leveraging an extended measure of Zero-Moment Point (ZMP)-driven rewards and task-driven rewards in a whole-body actor-critic framework, aiming to achieve coordinated actions of the upper and lower limbs for robust locomotion. Experiments conducted on a full-sized Unitree H1-2 robot verify the ability of our method to maintain balance on extremely narrow terrains and under external disturbances, demonstrating its effectiveness in enhancing the robot's adaptability to complex environments. The videos are given at https://whole-body-loco.github.io.