Live streaming platforms have become a dominant form of online content consumption, offering dynamically evolving content, real-time interactions, and highly engaging user experiences. These unique characteristics introduce new challenges that differentiate live streaming recommendation from traditional recommendation settings and have garnered increasing attention from industry in recent years. However, research progress in academia has been hindered by the lack of publicly available datasets that accurately reflect the dynamic nature of live streaming environments. To address this gap, we introduce KuaiLive, the first real-time, interactive dataset collected from Kuaishou, a leading live streaming platform in China with over 400 million daily active users. The dataset records the interaction logs of 23,772 users and 452,621 streamers over a 21-day period. Compared to existing datasets, KuaiLive offers several advantages: it includes precise live room start and end timestamps, multiple types of real-time user interactions (click, comment, like, gift), and rich side information features for both users and streamers. These features enable more realistic simulation of dynamic candidate items and better modeling of user and streamer behaviors. We conduct a thorough analysis of KuaiLive from multiple perspectives and evaluate several representative recommendation methods on it, establishing a strong benchmark for future research. KuaiLive can support a wide range of tasks in the live streaming domain, such as top-K recommendation, click-through rate prediction, watch time prediction, and gift price prediction. Moreover, its fine-grained behavioral data also enables research on multi-behavior modeling, multi-task learning, and fairness-aware recommendation. The dataset and related resources are publicly available at https://imgkkk574.github.io/KuaiLive.

Tool-Integrated Reasoning (TIR) empowers large language models (LLMs) to tackle complex tasks by interleaving reasoning steps with external tool interactions. However, existing reinforcement learning methods typically rely on outcome- or trajectory-level rewards, assigning uniform advantages to all steps within a trajectory. This coarse-grained credit assignment fails to distinguish effective tool calls from redundant or erroneous ones, particularly in long-horizon multi-turn scenarios. To address this, we propose MatchTIR, a framework that introduces fine-grained supervision via bipartite matching-based turn-level reward assignment and dual-level advantage estimation. Specifically, we formulate credit assignment as a bipartite matching problem between predicted and ground-truth traces, utilizing two assignment strategies to derive dense turn-level rewards. Furthermore, to balance local step precision with global task success, we introduce a dual-level advantage estimation scheme that integrates turn-level and trajectory-level signals, assigning distinct advantage values to individual interaction turns. Extensive experiments on three benchmarks demonstrate the superiority of MatchTIR. Notably, our 4B model surpasses the majority of 8B competitors, particularly in long-horizon and multi-turn tasks. Our codes are available at https://anonymous.4open.science/r/MatchTIR.

Tool-Integrated Reasoning (TIR) enables LLMs to solve complex tasks through iterative tool interactions. However, existing reinforcement learning methods often rely on trajectory-level supervision, limiting fine-grained credit assignment in long-horizon TIR scenarios. On-policy self-distillation offers denser signals through teacher branches with privileged context, but existing approaches typically derive such context from ground-truth answers or retrieved skills, which may not reflect the states actually visited by the agent. Moreover, token-level supervision fails to capture the turn-level structure of tool interactions. To address this, we propose TurnSight, a turn-level hindsight self-distillation framework that derives supervision directly from execution-conditioned hindsight. It then constructs multiple hindsight views with different lookahead horizons and selects reliable supervision through cross-horizon directional agreement. Finally, the selected hindsight signal is normalized across sibling rollouts and used to adaptively modulate RL advantages while preserving their original optimization direction. Extensive experiments on three benchmarks demonstrate the effectiveness of TurnSight. Our codes are available at https://github.com/quchangle1/TurnSight.

Information retrieval (IR) systems have traditionally been designed and trained for human users, with learning-to-rank methods relying heavily on large-scale human interaction logs such as clicks and dwell time. With the rapid emergence of large language model (LLM) powered search agents, however, retrieval is increasingly consumed by agents rather than human beings, and is embedded as a core component within multi-turn reasoning and action loops. In this setting, retrieval models trained under human-centric assumptions exhibit a fundamental mismatch with the way agents issue queries and consume results. In this work, we argue that retrieval models for agentic search should be trained directly from agent interaction data. We introduce learning to retrieve from agent trajectories as a new training paradigm, where supervision is derived from multi-step agent interactions. Through a systematic analysis of search agent trajectories, we identify key behavioral signals that reveal document utility, including browsing actions, unbrowsed rejections, and post-browse reasoning traces. Guided by these insights, we propose LRAT, a simple yet effective framework that mines high-quality retrieval supervision from agent trajectories and incorporates relevance intensity through weighted optimization. Extensive experiments on both in-domain and out-of-domain deep research benchmarks demonstrate that retrievers trained with LRAT consistently improve evidence recall, end-to-end task success, and execution efficiency across diverse agent architectures and scales. Our results highlight agent trajectories as a practical and scalable supervision source, pointing to a promising direction for retrieval in the era of agentic search.

Existing public live streaming datasets suffer from three major limitations: they provide limited access to temporally evolving multimodal live content, overlook users' cross-domain interactions between short videos and live streams, and contain only implicit behavioral signals without explicit feedback that captures users' perceived content quality and satisfaction. These limitations prevent existing benchmarks from faithfully reflecting real-world live streaming scenarios and hinder comprehensive research on live streaming recommendation. To address these limitations, we introduce KuaiLive-M3, a multi-modal, multi-domain, and multi-feedback dataset for live streaming recommendation, collected from Kuaishou, a leading live streaming and short video platform in China. KuaiLive-M3 covers 21,938 users and contains 35 million live streaming interactions and 111 million short video interactions, with fine-grained timestamps and diverse user behaviors. It further provides approximately 88 million timestamped segment-level multi-modal embeddings that capture the temporal evolution of live streaming content, as well as 25,403 questionnaire-based feedback records that bridge implicit user behaviors and explicit user preferences. Based on these unique signals, we establish benchmarks for cross-domain recommendation, live stream highlight prediction, and questionnaire-enhanced recommendation. Extensive experiments with representative baselines demonstrate that KuaiLive-M3 provides a challenging and realistic benchmark for future live streaming recommendation research. The results further highlight the importance of modeling temporally evolving content, transferring user preferences across domains, and bridging the gap between implicit behaviors and explicit user feedback. The dataset and benchmark code are publicly available at https://imgkkk574.github.io/KuaiLive-M3/.

Retrieval-augmented generation (RAG) has proven effective in enhancing the knowledge coverage of large language models (LLMs) and mitigating hallucinations by incorporating external retrieved documents. However, documents deemed relevant by the retriever are not necessarily helpful for answer generation, and including misleading information can even degrade performance. Existing efforts to estimate document utility often rely on the downstream generation performance, which conflates the influence of external documents with the intrinsic knowledge of the LLM, thereby obscuring the actual contribution of the retrieved content. To address this, this paper proposes Uplit-RAG, a uplift-driven knowledge preference alignment framework for RAG. Specifically, we first propose an uplift-based definition of document utility that quantifies each document’s marginal benefit over the LLM’s internal knowledge. We then optimize the reranker with three alignment objectives to identify and prioritize documents based on their uplift. This enables dynamic selection of documents that address the LLM’s knowledge gaps, going beyond fixed top-k selection, while reducing reference redundancy and the computational overhead of the LLM’s input. Extensive experiments demonstrate the effectiveness of Uplift-RAG.

Search clarification is a critical user interface for open-domain conversational Web search, where generating high-quality facets for ambiguous or multi-facet queries significantly guides disambiguation and enhances the user's interaction experience. Recently, in-context learning with Large Language Models (LLMs) has emerged as a promising approach for facet generation by leveraging static or similarity-based demonstrations as prompts. However, existing methods predominantly rely on query similarity, failing to account for the multi-dimensional nature of query intents. This limitation can lead LLMs to generate incorrect or suboptimal facets misaligned with user needs. To address this challenge, we propose an intent-covering framework that improves clarification facet generation by selecting demonstrations that comprehensively cover the diverse intents underlying a given query. Specifically, we first train a generative model with beam search to predict potential intents and construct an intent-document graph to capture their semantic relationships. We then introduce a heuristic greedy algorithm that optimizes demonstration selection by maximizing intent coverage. Furthermore, since the order of demonstrations significantly affects generation quality, we develop a re-ranking model to optimize their sequence for better contextual alignment. Experiments demonstrate the superiority of our approach over strong baselines in various lexical and semantic evaluation metrics. Additionally, we conduct an in-depth analysis of how the number, order, and contextual relevance of demonstrations influence generation performance.

Recently, live streaming services have seen a surge in popularity, prompting many platforms to offer both short video and live streaming services to meet the diverse needs of users and streamers. This has resulted in a close connection between short videos and live streaming within these platforms. Incorporating short video data into live streaming recommendation through cross-domain approaches can effectively mitigate the sparsity of live streaming gifting data. However, existing cross-domain recommendation methods primarily focus on transferring information across domains through overlapping users or items, while overlooking the strong connection between non-overlapping short videos and streamers. In this paper, we propose MGCCDR, a Multi-Graph Contrastive learning framework for Cross-Domain Recommendation, which leverages both overlapping users and non-overlapping items to enhance information transfer. Specifically, we first learn global representations from a global graph to establish connections between streamers and short videos. Subsequently, we construct three bipartite graphs among users, authors, and videos and introduce multi-graph learning to capture preferences within the target domain view, the source domain view, and the cross-domain view. Additionally, to address the varying contributions of each graph to the final recommendation task, we design an attention-based method to effectively integrate these representations, facilitating the information aggregation across domains. Extensive experiments on both commercial and public datasets demonstrate that our MGCCDR significantly outperforms the state-of-the-art methods.

Tool learning enables Large Language Models (LLMs) to interact with external environments by invoking tools, serving as an effective strategy to mitigate the limitations inherent in their pre-training data. In this process, tool documentation plays a crucial role by providing usage instructions for LLMs, thereby facilitating effective tool utilization. This paper concentrates on the critical challenge of bridging the comprehension gap between LLMs and external tools due to the inadequacies and inaccuracies inherent in existing human-centric tool documentation. We propose a novel framework, DRAFT, aimed at Dynamically Refining tool documentation through the Analysis of Feedback and Trials emanating from LLMs' interactions with external tools. This methodology pivots on an innovative trial-and-error approach, consisting of three distinct learning phases: experience gathering, learning from experience, and documentation rewriting, to iteratively enhance the tool documentation. This process is further optimized by implementing a diversity-promoting exploration strategy to ensure explorative diversity and a tool-adaptive termination mechanism to prevent overfitting while enhancing efficiency. Extensive experiments on multiple datasets demonstrate that DRAFT's iterative, feedback-based refinement significantly ameliorates documentation quality, fostering a deeper comprehension and more effective utilization of tools by LLMs. Notably, our analysis reveals that the tool documentation refined via our approach demonstrates robust cross-model generalization capabilities.

In this paper, we introduce the AI Search Paradigm, a comprehensive blueprint for next-generation search systems capable of emulating human information processing and decision-making. The paradigm employs a modular architecture of four LLM-powered agents (Master, Planner, Executor and Writer) that dynamically adapt to the full spectrum of information needs, from simple factual queries to complex multi-stage reasoning tasks. These agents collaborate dynamically through coordinated workflows to evaluate query complexity, decompose problems into executable plans, and orchestrate tool usage, task execution, and content synthesis. We systematically present key methodologies for realizing this paradigm, including task planning and tool integration, execution strategies, aligned and robust retrieval-augmented generation, and efficient LLM inference, spanning both algorithmic techniques and infrastructure-level optimizations. By providing an in-depth guide to these foundational components, this work aims to inform the development of trustworthy, adaptive, and scalable AI search systems.