Searcharxiv⌕ Search

arXiv subjects

Weitao Li

Publications and source records attributed to Weitao Li.

18 recordsLinked to original sources
arXiv2026

Enhancing LLM Metacognition via Cognitive Pairwise Training

Reinforcement learning with verifiable rewards (RLVR) has become central to LLM reasoning, but its outcome-level rewards can make models more willing to give confident answers when evidence or reasoning is unreliable. Existing SFT or RL methods mainly teach LLMs to refuse or express uncertainty at the response level, which can overfit abstention behavior rather than improve reasoning reliability. To address this limitation, we propose Cognitive Pairwise Training (CPT), a cognitive mid-training alignment stage that turns pairwise comparisons over reasoning traces into a reusable alignment signal. By learning to distinguish trustworthy from flawed reasoning, CPT encourages the model to internalize a reasoning-quality discrimination boundary rather than memorize surface refusal patterns. Across five model scales and three model families, CPT improves the reasoning--metacognition trade-off. At 14B, CPT+RL outperforms the standard SFT+RL pipeline by +2.2 math-average points and +5.6 abstention-F1 points. Further analyses show that CPT improves trace quality and exhibits strong robustness and scalability across evaluation and training settings. Code and models are released at https://github.com/Tsinghua-dhy/CPT.

cs.LG↗
arXiv2026

Generator-Aligned Representation Interfaces for Diagnostic Soft Equivariance

Exact-equivariant architectures typically encode prescribed group actions in specialized operators, which can complicate their reuse with generic backbones and across data modalities. We introduce the Generator-Aligned Representation Interface (GARI), a representation-level design principle that exposes selected transformation generators to a generic sequence backbone through aligned canonical and generator-induced views. We formalize the resulting behavior using a probe-specific soft-equivariance residual defined over declared data and transformation distributions. This framework distinguishes representation consistency from task robustness and exact equivariance, and localizes residual mismatch to interface construction, shared stream processing, and terminal fusion. We instantiate the interface as GARI-Net, which constructs generator-indexed streams, converts them into a common interaction frame, processes them with shared parameters, repairs ordering-induced context mismatch, enables cross-stream information exchange, and aggregates them using inter-stream discrepancy. Direct Equivariance Error (DEE) provides a frozen-checkpoint diagnostic of the prescribed representation relation under known token or voxel actions. Experiments on genomic sequences, images, and three-dimensional point clouds examine sequence reversal, planar rotations and reflections, and controlled axial transfer. Across these settings, the same interface principle supports task-relevant transformation consistency and generalization to declared held-out probes without requiring group-specific redesign of the sequence backbone. GARI therefore provides a portable diagnostic complement to hard-equivariant architectures: it makes generator structure accessible, learnable, and measurable, while finite-probe evidence remains distinct from certification of exact equivariance over a continuous group.

cs.LG↗
arXiv2026

ClawRec: A Claw-Native Recommender System

Recommender systems have become integral to navigating the modern digital ecosystem. Yet most deployed systems remain confined within single-platform boundaries, observing localized interaction traces and ranking items from isolated candidate spaces. This design is poorly suited to real-world tasks that unfold through searches, content consumption, and comparisons across multiple information sources. Claw-style personal agents, with persistent access to authorized cross-platform context, create an opportunity for recommendation to operate around the user rather than any single platform. In this paper, we introduce Claw-native recommender systems, a new paradigm that moves beyond platform-local ranking to produce unified, complementary recommendation slates spanning diverse sources and content forms. To instantiate this paradigm, we present ClawRec, the first recommender system designed to operate natively in this environment. ClawRec maintains an evidence-linked, temporally structured user state that connects cross-platform behaviors with cross-source recommendations. It organizes retrieval around functional source roles and selects candidates according to their marginal utility, producing non-redundant slates aligned with the user's active task. To enable rigorous evaluation, we introduce ClawRec-SimBench, a benchmark constructed from sequences of concrete life events and cross-platform behavior trajectories. Experiments show that ClawRec outperforms the strongest baselines, achieving an NDCG@20 of 0.6134 (+0.1126) and a Hit@20 of 0.6944 (+0.0854), while also improving user state quality and temporal alignment. Our code and dataset are available at https://github.com/RUCAIBox/ClawRec.

cs.IR↗
arXiv2026

UR$^2$: Unify RAG and Reasoning through Reinforcement Learning

Large Language Models (LLMs) have shown strong capabilities through two complementary paradigms: Retrieval-Augmented Generation (RAG) for knowledge grounding and Reinforcement Learning from Verifiable Rewards (RLVR) for complex reasoning. However, existing attempts to unify these paradigms remain narrow in scope, typically limited to open-domain QA with fixed retrieval settings, which constrains generalization to broader domains. To address this limitation, we propose UR$^2$ (Unified RAG and Reasoning)), a general reinforcement learning framework that dynamically coordinates retrieval and reasoning. UR$^2$ introduces two key designs: a difficulty-aware curriculum that selectively invokes retrieval only for challenging instances, and a hybrid knowledge access strategy that combines domain-specific offline corpora with on-the-fly LLM-generated summaries. Together, these components mitigate the imbalance between retrieval and reasoning and improve robustness to noisy information. Experiments on open-domain QA, MMLU-Pro, medical, and mathematical reasoning tasks show that UR$^2$, built on Qwen-2.5-3/7B and LLaMA-3.1-8B, consistently outperforms existing RAG and RL baselines, and achieves performance comparable to GPT-4o-mini and GPT-4.1-mini on several benchmarks. Our code is available at https://github.com/Tsinghua-dhy/UR2.

cs.CL↗
arXiv2026

Beyond "I Don't Know": Evaluating LLM Self-Awareness in Discriminating Data and Model Uncertainty

Reliable Large Language Models (LLMs) should abstain when confidence is insufficient. However, prior studies often treat refusal as a generic "I don't know'', failing to distinguish input-level ambiguity (data uncertainty) from capability limitations (model uncertainty). This lack of distinction limits downstream action decisions like requesting clarification or invoking external tools. In this work, we introduce UA-Bench, a benchmark of over 3,500 questions drawn from six datasets spanning knowledge-intensive and reasoning-intensive tasks, designed to evaluate explicit uncertainty attribution. An evaluation of 18 frontier LLMs shows that even state-of-the-art models struggle to reliably discriminate between data uncertainty and model uncertainty, and that high answer accuracy does not necessarily imply strong uncertainty attribution ability. To narrow this gap, we propose a lightweight data synthesis and reinforcement learning strategy. Experiments on both Qwen3-4B-Instruct-2507 and Qwen3-8B in thinking mode show that the proposed method improves uncertainty attribution while preserving answer accuracy. Our code and data are publicly available now.

cs.CL↗
arXiv2026

Deep Research for Recommender Systems

The technical foundations of recommender systems have progressed from collaborative filtering to complex neural models and, more recently, large language models. Despite these technological advances, deployed systems often underserve their users by simply presenting a list of items, leaving the burden of exploration, comparison, and synthesis entirely on the user. This paper argues that this traditional "tool-based" paradigm fundamentally limits user experience, as the system acts as a passive filter rather than an active assistant. To address this limitation, we propose a novel deep research paradigm for recommendation, which replaces conventional item lists with comprehensive, user-centric reports. We instantiate this paradigm through RecPilot, a multi-agent framework comprising two core components: a user trajectory simulation agent that autonomously explores the item space, and a self-evolving report generation agent that synthesizes the findings into a coherent, interpretable report tailored to support user decisions. This approach reframes recommendation as a proactive, agent-driven service. Extensive experiments on public datasets demonstrate that RecPilot not only achieves strong performance in modeling user behaviors but also generates highly persuasive reports that substantially reduce user effort in item evaluation, validating the potential of this new interaction paradigm.

cs.IR↗
arXiv2026

Towards Transparent RAG: Fostering Evidence Traceability in LLM Generation via Reinforcement Learning

Retrieval-Augmented Generation (RAG) delivers substantial value in knowledge-intensive applications. However, its generated responses often lack transparent reasoning paths that trace back to source evidence from retrieved documents. This opacity not only compromises the interpretability of the output but also limits the model's ability to fully exploit the provided context. To address this, we propose TRACE (Transparent RAG with evidenCE tracing), a framework designed to enhance evidence traceability in Large Language Models (LLMs) through reinforcement learning (RL). TRACE guides LLMs to produce structured outputs with explicit evidence citations by prompting and rewarding evidence relevance and proper formatting, alongside accuracy, to optimize structured traceability. To ensure training stability with multiple reward signals, we further introduce an adaptive strategy for merging rewards and adopt a stabilized KL-divergence estimator. Experiments on three multi-hop QA datasets using Qwen2.5-7B-Instruct and Llama-3.1-8B-Instruct show that TRACE achieves both transparent, evidence-attributed outputs and accuracy improvements of 10-30%. The resulting performance is comparable to advanced commercial LLMs (e.g., OpenAI o1, DeepSeek-R1). Further analyses demonstrate strong generalization capabilities to unseen tasks. Our code is publicly available now.

cs.CL↗
arXiv2026

RecNet: Self-Evolving Preference Propagation for Agentic Recommender Systems

Agentic recommender systems leverage Large Language Models (LLMs) to model complex user behaviors and support personalized decision-making. However, existing methods primarily model preference changes based on explicit user-item interactions, which are sparse, noisy, and unable to reflect the real-time, mutual influences among users and items. To address these limitations, we propose RecNet, a self-evolving preference propagation framework that proactively propagates real-time preference updates across related users and items. RecNet consists of two complementary phases. In the forward phase, the centralized preference routing mechanism leverages router agents to integrate preference updates and dynamically propagate them to the most relevant agents. To ensure accurate and personalized integration of propagated preferences, we further introduce a personalized preference reception mechanism, which combines a message buffer for temporary caching and an optimizable, rule-based filter memory to guide selective preference assimilation based on past experience and interests. In the backward phase, the feedback-driven propagation optimization mechanism simulates a multi-agent reinforcement learning framework, using LLMs for credit assignment, gradient analysis, and module-level optimization, enabling continuous self-evolution of propagation strategies. Extensive experiments on various scenarios demonstrate the effectiveness of RecNet in modeling preference propagation for recommender systems.

cs.AI↗
arXiv2025

Efficient Dynamic Clustering-Based Document Compression for Retrieval-Augmented-Generation

Retrieval-Augmented Generation (RAG) has emerged as a widely adopted approach for knowledge injection during large language model (LLM) inference in recent years. However, due to their limited ability to exploit fine-grained inter-document relationships, current RAG implementations face challenges in effectively addressing the retrieved noise and redundancy content, which may cause error in the generation results. To address these limitations, we propose an Efficient Dynamic Clustering-based document Compression framework (EDC2-RAG) that utilizes latent inter-document relationships while simultaneously removing irrelevant information and redundant content. We validate our approach, built upon GPT-3.5-Turbo and GPT-4o-mini, on widely used knowledge-QA and Hallucination-Detection datasets. Experimental results show that our method achieves consistent performance improvements across various scenarios and experimental settings, demonstrating strong robustness and applicability. Our code and datasets are available at https://github.com/Tsinghua-dhy/EDC-2-RAG.

cs.CL↗
arXiv2025

iFairy: the First 2-bit Complex LLM with All Parameters in $\{\pm1, \pm i\}$

Quantization-Aware Training (QAT) integrates quantization into the training loop, enabling LLMs to learn robust low-bit representations, and is widely recognized as one of the most promising research directions. All current QAT research focuses on minimizing quantization error on full-precision models, where the full-precision accuracy acts as an upper bound (accuracy ceiling). No existing method has even attempted to surpass this ceiling. To break this ceiling, we propose a new paradigm: raising the ceiling (full-precision model), and then still quantizing it efficiently into 2 bits. We propose Fairy$\pm i$, the first 2-bit quantization framework for complex-valued LLMs. Specifically, our method leverages the representational advantages of the complex domain to boost full-precision accuracy. We map weights to the fourth roots of unity $\{\pm1, \pm i\}$, forming a perfectly symmetric and information-theoretically optimal 2-bit representation. Importantly, each quantized weight has either a zero real or imaginary part, enabling multiplication-free inference using only additions and element swaps. Experimental results show that Fairy$\pm i$ outperforms the ceiling of existing 2-bit quantization approaches in terms of both PPL and downstream tasks, while maintaining strict storage and compute efficiency. This work opens a new direction for building highly accurate and practical LLMs under extremely low-bit constraints.

cs.LG↗
arXiv2025

Citation-Enhanced Generation for LLM-based Chatbots

Large language models (LLMs) exhibit powerful general intelligence across diverse scenarios, including their integration into chatbots. However, a vital challenge of LLM-based chatbots is that they may produce hallucinated content in responses, which significantly limits their applicability. Various efforts have been made to alleviate hallucination, such as retrieval augmented generation and reinforcement learning with human feedback, but most of them require additional training and data annotation. In this paper, we propose a novel post-hoc Citation-Enhanced Generation (CEG) approach combined with retrieval argumentation. Unlike previous studies that focus on preventing hallucinations during generation, our method addresses this issue in a post-hoc way. It incorporates a retrieval module to search for supporting documents relevant to the generated content, and employs a natural language inference-based citation generation module. Once the statements in the generated content lack of reference, our model can regenerate responses until all statements are supported by citations. Note that our method is a training-free plug-and-play plugin that is capable of various LLMs. Experiments on various hallucination-related datasets show our framework outperforms state-of-the-art methods in both hallucination detection and response regeneration on three benchmarks. Our codes and dataset will be publicly available.

cs.CL↗
arXiv2025

Agent Hospital: A Simulacrum of Hospital with Evolvable Medical Agents

The recent rapid development of large language models (LLMs) has sparked a new wave of technological revolution in medical artificial intelligence (AI). While LLMs are designed to understand and generate text like a human, autonomous agents that utilize LLMs as their "brain" have exhibited capabilities beyond text processing such as planning, reflection, and using tools by enabling their "bodies" to interact with the environment. We introduce a simulacrum of hospital called Agent Hospital that simulates the entire process of treating illness, in which all patients, nurses, and doctors are LLM-powered autonomous agents. Within the simulacrum, doctor agents are able to evolve by treating a large number of patient agents without the need to label training data manually. After treating tens of thousands of patient agents in the simulacrum (human doctors may take several years in the real world), the evolved doctor agents outperform state-of-the-art medical agent methods on the MedQA benchmark comprising US Medical Licensing Examination (USMLE) test questions. Our methods of simulacrum construction and agent evolution have the potential in benefiting a broad range of applications beyond medical AI.

cs.AI↗
arXiv2024

EXIT: An EXplicit Interest Transfer Framework for Cross-Domain Recommendation

Cross-domain recommendation has attracted substantial interest in industrial apps such as Meituan, which serves multiple business domains via knowledge transfer and meets the diverse interests of users. However, existing methods typically follow an implicit modeling paradigm that blends the knowledge from both the source and target domains, and design intricate network structures to share learned embeddings or patterns between domains to improve recommendation accuracy. Since the transfer of interest signals is unsupervised, these implicit paradigms often struggle with the negative transfer resulting from differences in service functions and presentation forms across different domains. In this paper, we propose a simple and effective EXplicit Interest Transfer framework named EXIT to address the stated challenge. Specifically, we propose a novel label combination approach that enables the model to directly learn beneficial source domain interests through supervised learning, while excluding inappropriate interest signals. Moreover, we introduce a scene selector network to model the interest transfer intensity under fine-grained scenes. Offline experiments conducted on the industrial production dataset and online A/B tests validate the superiority and effectiveness of our proposed framework. Without complex network structures or training processes, EXIT can be easily deployed in the industrial recommendation system. EXIT has been successfully deployed in the online homepage recommendation system of Meituan App, serving the main traffic.

cs.IR↗
arXiv2024

Reinforcement Learning Based Robust Volt/Var Control in Active Distribution Networks With Imprecisely Known Delay

Active distribution networks (ADNs) incorporating massive photovoltaic (PV) devices encounter challenges of rapid voltage fluctuations and potential violations. Due to the fluctuation and intermittency of PV generation, the state gap, arising from time-inconsistent states and exacerbated by imprecisely known system delays, significantly impacts the accuracy of voltage control. This paper addresses this challenge by introducing a framework for delay adaptive Volt/Var control (VVC) in the presence of imprecisely known system delays to regulate the reactive power of PV inverters. The proposed approach formulates the voltage control, based on predicted system operation states, as a robust VVC problem. It employs sample selection from the state prediction interval to promptly identify the worst-performing system operation state. Furthermore, we leverage the decentralized partially observable Markov decision process (Dec-POMDP) to reformulate the robust VVC problem. We design Multiple Policy Networks and employ Multiple Policy Networks and Reward Shaping-based Multi-agent Twin Delayed Deep Deterministic Policy Gradient (MPNRS-MATD3) algorithm to efficiently address and solve the Dec-POMDP model-based problem. Simulation results show the delay adaption characteristic of our proposed framework, and the MPNRS-MATD3 outperforms other multi-agent reinforcement learning algorithms in robust voltage control.

eess.SY↗
arXiv2023

Incipient Fault Detection in Power Distribution System: A Time-Frequency Embedded Deep Learning Based Approach

Incipient fault detection in power distribution systems is crucial to improve the reliability of the grid. However, the non-stationary nature and the inadequacy of the training dataset due to the self-recovery of the incipient fault signal, make the incipient fault detection in power distribution systems a great challenge. In this paper, we focus on incipient fault detection in power distribution systems and address the above challenges. In particular, we propose an ADaptive Time-Frequency Memory(AD-TFM) cell by embedding wavelet transform into the Long Short-Term Memory (LSTM), to extract features in time and frequency domain from the non-stationary incipient fault signals.We make scale parameters and translation parameters of wavelet transform learnable to adapt to the dynamic input signals. Based on the stacked AD-TFM cells, we design a recurrent neural network with ATtention mechanism, named AD-TFM-AT model, to detect incipient fault with multi-resolution and multi-dimension analysis. In addition, we propose two data augmentation methods, namely phase switching and temporal sliding, to effectively enlarge the training datasets. Experimental results on two open datasets show that our proposed AD-TFM-AT model and data augmentation methods achieve state-of-the-art (SOTA) performance of incipient fault detection in power distribution system. We also disclose one used dataset logged at State Grid Corporation of China to facilitate future research.

eess.SP↗
arXiv2022

Autonomous Smart Grid Fault Detection

Smart grid plays a crucial role for the smart society and the upcoming carbon neutral society. Achieving autonomous smart grid fault detection is critical for smart grid system state awareness, maintenance and operation. This paper focuses on fault monitoring in smart grid and discusses the inherent technical challenges and solutions. In particular, we first present the basic principles of smart grid fault detection. Then, we explain the new requirements for autonomous smart grid fault detection, the technical challenges and their possible solutions. A case study is introduced, as a preliminary study for autonomous smart grid fault detection. In addition, we highlight relevant directions for future research.

cs.DC↗
arXiv2020

Throughput Optimal Uplink Scheduling in Heterogeneous PLC and LTE Communication for Delay Aware Smart Grid Applications

Smart grid is an energy network that integrates advanced power equipment, communication technology and control technology. It can transmit two-way power and data among all components of the grid at the same time. The existing smart grid communication technologies include power line carrier (PLC) com-munication, industrial Ethernet, passive optical networks and wireless communi-cation, each of which have different advantages. Due to the complex application scenarios, massive sampling points and high transmission reliability require-ments, a single communication method cannot fully meet the communication re-quirements of smart grid, and heterogeneous communication modes are required. In addition, with the development of cellular technology, long term evolution (LTE)-based standards have been identified as a promising technology that can meet the strict requirements of various operations in smart grid. In this paper, we analyze the advantages and disadvantages of PLC and LTE communication, and design a network framework for PLC and LTE communication uplink heteroge-neous communication in smart grid. Then, we propose an uplink scheduling transmission method for sampling data with optimized throughput according to the requirements of system delay and reliability. Then, we use the formula deri-vation to prove the stability and solvability of the scheduling system in theory. Finally, the simulation results show that under the condition of satisfying the de-lay requirement, our proposed framework can optimally allocate the wireless communication resource and maximize the throughput of the uplink transmission system. Keywords: smart grid, heterogeneous communication, optimized throughput.

cs.NI↗
arXiv2020

TIMELY: Pushing Data Movements and Interfaces in PIM Accelerators Towards Local and in Time Domain

Resistive-random-access-memory (ReRAM) based processing-in-memory (R$^2$PIM) accelerators show promise in bridging the gap between Internet of Thing devices' constrained resources and Convolutional/Deep Neural Networks' (CNNs/DNNs') prohibitive energy cost. Specifically, R$^2$PIM accelerators enhance energy efficiency by eliminating the cost of weight movements and improving the computational density through ReRAM's high density. However, the energy efficiency is still limited by the dominant energy cost of input and partial sum (Psum) movements and the cost of digital-to-analog (D/A) and analog-to-digital (A/D) interfaces. In this work, we identify three energy-saving opportunities in R$^2$PIM accelerators: analog data locality, time-domain interfacing, and input access reduction, and propose an innovative R$^2$PIM accelerator called TIMELY, with three key contributions: (1) TIMELY adopts analog local buffers (ALBs) within ReRAM crossbars to greatly enhance the data locality, minimizing the energy overheads of both input and Psum movements; (2) TIMELY largely reduces the energy of each single D/A (and A/D) conversion and the total number of conversions by using time-domain interfaces (TDIs) and the employed ALBs, respectively; (3) we develop an only-once input read (O$^2$IR) mapping method to further decrease the energy of input accesses and the number of D/A conversions. The evaluation with more than 10 CNN/DNN models and various chip configurations shows that, TIMELY outperforms the baseline R$^2$PIM accelerator, PRIME, by one order of magnitude in energy efficiency while maintaining better computational density (up to 31.2$\times$) and throughput (up to 736.6$\times$). Furthermore, comprehensive studies are performed to evaluate the effectiveness of the proposed ALB, TDI, and O$^2$IR innovations in terms of energy savings and area reduction.

cs.LG↗