arXiv subjects

Qi Zhang

Publications and source records attributed to Qi Zhang.

At least 433 records · Page 24Linked to original sources

arXiv2025

Improving Prediction Certainty Estimation for Reliable Early Exiting via Null Space Projection

Early exiting has demonstrated great potential in accelerating the inference of pre-trained language models (PLMs) by enabling easy samples to exit at shallow layers, eliminating the need for executing deeper layers. However, existing early exiting methods primarily rely on class-relevant logits to formulate their exiting signals for estimating prediction certainty, neglecting the detrimental influence of class-irrelevant information in the features on prediction certainty. This leads to an overestimation of prediction certainty, causing premature exiting of samples with incorrect early predictions. To remedy this, we define an NSP score to estimate prediction certainty by considering the proportion of class-irrelevant information in the features. On this basis, we propose a novel early exiting method based on the Certainty-Aware Probability (CAP) score, which integrates insights from both logits and the NSP score to enhance prediction certainty estimation, thus enabling more reliable exiting decisions. The experimental results on the GLUE benchmark show that our method can achieve an average speed-up ratio of 2.19x across all tasks with negligible performance degradation, surpassing the state-of-the-art (SOTA) ConsistentEE by 28%, yielding a better trade-off between task performance and inference efficiency. The code is available at https://github.com/He-Jianing/NSP.git.

cs.LG↗

arXiv2025

Self-Evolved Reward Learning for LLMs

Reinforcement Learning from Human Feedback (RLHF) is a crucial technique for aligning language models with human preferences, playing a pivotal role in the success of conversational models like GPT-4, ChatGPT, and Llama 2. A core challenge in employing RLHF lies in training a reliable reward model (RM), which relies on high-quality labels typically provided by human experts or advanced AI system. These methods can be costly and may introduce biases that affect the language model's responses. As language models improve, human input may become less effective in further enhancing their performance. In this paper, we propose Self-Evolved Reward Learning (SER), a novel approach where the RM generates additional training data to iteratively improve itself. We conducted extensive experiments on multiple datasets such as HH-RLHF and UltraFeedback, using models like Mistral and Llama 3, and compare SER against various baselines. Our results demonstrate that even with limited human-annotated data, learning from self-feedback can robustly enhance RM performance, thereby boosting the capabilities of large language models (LLMs). Resources of this paper can be found at https://aka.ms/ser

cs.CL↗

arXiv2025

RACE-Align: Retrieval-Augmented and Chain-of-Thought Enhanced Preference Alignment for Large Language Models

Large Language Models (LLMs) struggle with accuracy, domain-specific reasoning, and interpretability in vertical domains. Traditional preference alignment methods like Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimization (DPO) often overlook the underlying knowledge sources and reasoning logic. This paper introduces RACE-Align (Retrieval-Augmented and Chain-of-Thought Enhanced Alignment), a novel framework designed to address these limitations. RACE-Align systematically constructs a binary preference dataset incorporating external knowledge support and explicit Chain-of-Thought (CoT) reasoning, then aligns LLMs using the DPO algorithm. The core innovation lies in its preference data construction strategy: it integrates AI-driven retrieval for factual grounding, enhancing knowledgeability and accuracy, and emphasizes the optimization of domain-specific CoT, treating the reasoning process itself as a key preference dimension. A multi-stage, AI-driven refinement pipeline cost-effectively generates these preference pairs. Experimental validation in Traditional Chinese Medicine (TCM) using Qwen3-1.7B as the base model demonstrates that RACE-Align significantly outperforms the original base model and a model fine-tuned only with Supervised Fine-Tuning (SFT). Improvements were observed across multiple dimensions, including answer accuracy, information richness, application of TCM thinking patterns, logicality and depth of reasoning, and interpretability. These findings suggest RACE-Align offers an effective pathway to enhance LLMs' knowledge application, reasoning reliability, and process transparency in complex vertical domains.

cs.CL↗

arXiv2025

Measuring Data Diversity for Instruction Tuning: A Systematic Analysis and A Reliable Metric

Data diversity is crucial for the instruction tuning of large language models. Existing studies have explored various diversity-aware data selection methods to construct high-quality datasets and enhance model performance. However, the fundamental problem of precisely defining and measuring data diversity remains underexplored, limiting clear guidance for data engineering. To address this, we systematically analyze 11 existing diversity measurement methods by evaluating their correlation with model performance through extensive fine-tuning experiments. Our results indicate that a reliable diversity measure should properly account for both inter-sample differences and the information density in the sample space. Building on this, we propose NovelSum, a new diversity metric based on sample-level "novelty." Experiments on both simulated and real-world data show that NovelSum accurately captures diversity variations and achieves a 0.97 correlation with instruction-tuned model performance, highlighting its value in guiding data engineering practices. With NovelSum as an optimization objective, we further develop a greedy, diversity-oriented data selection strategy that outperforms existing approaches, validating both the effectiveness and practical significance of our metric. The code is available at https://github.com/UmeanNever/NovelSum.

cs.CL↗

arXiv2025

TSD-SR: One-Step Diffusion with Target Score Distillation for Real-World Image Super-Resolution

Pre-trained text-to-image diffusion models are increasingly applied to real-world image super-resolution (Real-ISR) task. Given the iterative refinement nature of diffusion models, most existing approaches are computationally expensive. While methods such as SinSR and OSEDiff have emerged to condense inference steps via distillation, their performance in image restoration or details recovery is not satisfied. To address this, we propose TSD-SR, a novel distillation framework specifically designed for real-world image super-resolution, aiming to construct an efficient and effective one-step model. We first introduce the Target Score Distillation, which leverages the priors of diffusion models and real image references to achieve more realistic image restoration. Secondly, we propose a Distribution-Aware Sampling Module to make detail-oriented gradients more readily accessible, addressing the challenge of recovering fine details. Extensive experiments demonstrate that our TSD-SR has superior restoration results (most of the metrics perform the best) and the fastest inference speed (e.g. 40 times faster than SeeSR) compared to the past Real-ISR approaches based on pre-trained diffusion priors.

cs.CV↗

arXiv2025

GeAR: Generation Augmented Retrieval

Document retrieval techniques are essential for developing large-scale information systems. The common approach involves using a bi-encoder to compute the semantic similarity between a query and documents. However, the scalar similarity often fail to reflect enough information, hindering the interpretation of retrieval results. In addition, this process primarily focuses on global semantics, overlooking the finer-grained semantic relationships between the query and the document's content. In this paper, we introduce a novel method, $\textbf{Ge}$neration $\textbf{A}$ugmented $\textbf{R}$etrieval ($\textbf{GeAR}$), which not only improves the global document-query similarity through contrastive learning, but also integrates well-designed fusion and decoding modules. This enables GeAR to generate relevant context within the documents based on a given query, facilitating learning to retrieve local fine-grained information. Furthermore, when used as a retriever, GeAR does not incur any additional computational cost over bi-encoders. GeAR exhibits competitive retrieval performance across diverse scenarios and tasks. Moreover, qualitative analysis and the results generated by GeAR provide novel insights into the interpretation of retrieval results. The code, data, and models will be released at \href{https://github.com/microsoft/LMOps}{https://github.com/microsoft/LMOps}.

cs.IR↗

arXiv2025

The First MPDD Challenge: Multimodal Personality-aware Depression Detection

Depression is a widespread mental health issue affecting diverse age groups, with notable prevalence among college students and the elderly. However, existing datasets and detection methods primarily focus on young adults, neglecting the broader age spectrum and individual differences that influence depression manifestation. Current approaches often establish a direct mapping between multimodal data and depression indicators, failing to capture the complexity and diversity of depression across individuals. This challenge includes two tracks based on age-specific subsets: Track 1 uses the MPDD-Elderly dataset for detecting depression in older adults, and Track 2 uses the MPDD-Young dataset for detecting depression in younger participants. The Multimodal Personality-aware Depression Detection (MPDD) Challenge aims to address this gap by incorporating multimodal data alongside individual difference factors. We provide a baseline model that fuses audio and video modalities with individual difference information to detect depression manifestations in diverse populations. This challenge aims to promote the development of more personalized and accurate de pression detection methods, advancing mental health research and fostering inclusive detection systems. More details are available on the official challenge website: https://hacilab.github.io/MPDDChallenge.github.io.

cs.AI↗

arXiv2025

Neural Canonical Transformations for Quantum Anharmonic Solids of Lithium

Lithium is a typical quantum solid, characterized by cubic structures at ambient pressure. As the pressure increases, it forms more complex structures and undergoes a metal-to-semiconductor transformation, complicating theoretical and experimental analyses. We employ the neural canonical transformation approach, an \textit{ab initio} variational method based on probabilistic generative models, to investigate the quantum anharmonic effects in lithium solids at finite temperatures. This approach combines a normalizing flow for phonon excited-state wave functions with a probabilistic model for the occupation of energy levels, optimized jointly to minimize the free energy. Our results indicate that quantum anharmonicity lowers the \textit{bcc}-\textit{fcc} transition temperature compared to classical molecular dynamics predictions. At high pressures, the predicted fractional coordinates of lithium atoms in the \textit{cI16} structure show good quantitative agreement with experimental observations. Finally, contrary to previous beliefs, we find that the poor metallic \textit{oC88} structure is stabilized by the potential energy surface obtained via high-accuracy electronic structure calculations, rather than thermal or quantum nuclear effects.

cond-mat.mtrl-sci↗

arXiv2025

A Comprehensive Survey of Artificial Intelligence Techniques for Talent Analytics

In today's competitive and fast-evolving business environment, it is a critical time for organizations to rethink how to make talent-related decisions in a quantitative manner. Indeed, the recent development of Big Data and Artificial Intelligence (AI) techniques have revolutionized human resource management. The availability of large-scale talent and management-related data provides unparalleled opportunities for business leaders to comprehend organizational behaviors and gain tangible knowledge from a data science perspective, which in turn delivers intelligence for real-time decision-making and effective talent management at work for their organizations. In the last decade, talent analytics has emerged as a promising field in applied data science for human resource management, garnering significant attention from AI communities and inspiring numerous research efforts. To this end, we present an up-to-date and comprehensive survey on AI technologies used for talent analytics in the field of human resource management. Specifically, we first provide the background knowledge of talent analytics and categorize various pertinent data. Subsequently, we offer a comprehensive taxonomy of relevant research efforts, categorized based on three distinct application-driven scenarios: talent management, organization management, and labor market analysis. In conclusion, we summarize the open challenges and potential prospects for future research directions in the domain of AI-driven talent analytics.

cs.CY↗

arXiv2025

LeTS: Learning to Think-and-Search via Process-and-Outcome Reward Hybridization

Large language models (LLMs) have demonstrated impressive capabilities in reasoning with the emergence of reasoning models like OpenAI-o1 and DeepSeek-R1. Recent research focuses on integrating reasoning capabilities into the realm of retrieval-augmented generation (RAG) via outcome-supervised reinforcement learning (RL) approaches, while the correctness of intermediate think-and-search steps is usually neglected. To address this issue, we design a process-level reward module to mitigate the unawareness of intermediate reasoning steps in outcome-level supervision without additional annotation. Grounded on this, we propose Learning to Think-and-Search (LeTS), a novel framework that hybridizes stepwise process reward and outcome-based reward to current RL methods for RAG. Extensive experiments demonstrate the generalization and inference efficiency of LeTS across various RAG benchmarks. In addition, these results reveal the potential of process- and outcome-level reward hybridization in boosting LLMs' reasoning ability via RL under other scenarios. The code will be released soon.

cs.CL↗

arXiv2025

RePrompt: Reasoning-Augmented Reprompting for Text-to-Image Generation via Reinforcement Learning

Despite recent progress in text-to-image (T2I) generation, existing models often struggle to faithfully capture user intentions from short and under-specified prompts. While prior work has attempted to enhance prompts using large language models (LLMs), these methods frequently generate stylistic or unrealistic content due to insufficient grounding in visual semantics and real-world composition. Inspired by recent advances in reasoning for language model, we propose RePrompt, a novel reprompting framework that introduces explicit reasoning into the prompt enhancement process via reinforcement learning. Instead of relying on handcrafted rules or stylistic rewrites, our method trains a language model to generate structured, self-reflective prompts by optimizing for image-level outcomes. The tailored reward models assesse the generated images in terms of human preference, semantic alignment, and visual composition, providing indirect supervision to refine prompt generation. Our approach enables end-to-end training without human-annotated data. Experiments on GenEval and T2I-Compbench show that RePrompt significantly boosts spatial layout fidelity and compositional generalization across diverse T2I backbones, establishing new state-of-the-art results.

cs.CV↗

arXiv2025

Quantum melting of generalized electron crystal in twisted bilayer MoSe2

Electrons can form an ordered solid crystal phase ascribed to the interplay between Coulomb repulsion and kinetic energy. Tuning these energy scales can drive a phase transition from electron solid to liquid, i.e. melting of Wigner crystal. Generalized Wigner crystals (GWCs) pinned to moire superlattices have been reported by optical and scanning-probe-based methods. Using transport measurements to investigate GWCs is vital to a complete characterization, however, still poses a significant challenge due to difficulties in making reliable electrical contacts. Here, we report the electrical transport detection of GWCs at fractional fillings nu = 2/5, 1/2, 3/5, 2/3, 8/9, 10/9, and 4/3 in twisted bilayer MoSe2. We further observe that these GWCs undergo continuous quantum melting transitions to liquid phases by tuning doping density, magnetic and displacement fields, manifested by quantum critical scaling behaviors. Our findings establish twisted bilayer MoSe2 as a novel system to study strongly correlated states of matter and their quantum phase transitions.

cond-mat.str-el↗

arXiv2025

DocFusion: A Unified Framework for Document Parsing Tasks

Document parsing is essential for analyzing complex document structures and extracting fine-grained information, supporting numerous downstream applications. However, existing methods often require integrating multiple independent models to handle various parsing tasks, leading to high complexity and maintenance overhead. To address this, we propose DocFusion, a lightweight generative model with only 0.28B parameters. It unifies task representations and achieves collaborative training through an improved objective function. Experiments reveal and leverage the mutually beneficial interaction among recognition tasks, and integrating recognition data significantly enhances detection performance. The final results demonstrate that DocFusion achieves state-of-the-art (SOTA) performance across four key tasks.

cs.CL↗

arXiv2025

When LLMs meet open-world graph learning: a new perspective for unlabeled data uncertainty

Recently, large language models (LLMs) have significantly advanced text-attributed graph (TAG) learning. However, existing methods inadequately handle data uncertainty in open-world scenarios, especially concerning limited labeling and unknown-class nodes. Prior solutions typically rely on isolated semantic or structural approaches for unknown-class rejection, lacking effective annotation pipelines. To address these limitations, we propose Open-world Graph Assistant (OGA), an LLM-based framework that combines adaptive label traceability, which integrates semantics and topology for unknown-class rejection, and a graph label annotator to enable model updates using newly annotated nodes. Comprehensive experiments demonstrate OGA's effectiveness and practicality.

cs.LG↗

arXiv2025

ToolHop: A Query-Driven Benchmark for Evaluating Large Language Models in Multi-Hop Tool Use

Effective evaluation of multi-hop tool use is critical for analyzing the understanding, reasoning, and function-calling capabilities of large language models (LLMs). However, progress has been hindered by a lack of reliable evaluation datasets. To address this, we present ToolHop, a dataset comprising 995 user queries and 3,912 associated tools, specifically designed for rigorous evaluation of multi-hop tool use. ToolHop ensures diverse queries, meaningful interdependencies, locally executable tools, detailed feedback, and verifiable answers through a novel query-driven data construction approach that includes tool creation, document refinement, and code generation. We evaluate 14 LLMs across five model families (i.e., LLaMA3.1, Qwen2.5, Gemini1.5, Claude3.5, and GPT), uncovering significant challenges in handling multi-hop tool-use scenarios. The leading model, GPT-4o, achieves an accuracy of 49.04%, underscoring substantial room for improvement. Further analysis reveals variations in tool-use strategies for various families, offering actionable insights to guide the development of more effective approaches. Code and data can be found in https://huggingface.co/datasets/bytedance-research/ToolHop.

cs.CL↗

arXiv2025

AXIS: Efficient Human-Agent-Computer Interaction with API-First LLM-Based Agents

Multimodal large language models (MLLMs) have enabled LLM-based agents to directly interact with application user interfaces (UIs), enhancing agents' performance in complex tasks. However, these agents often suffer from high latency and low reliability due to the extensive sequential UI interactions. To address this issue, we propose AXIS, a novel LLM-based agents framework that prioritize actions through application programming interfaces (APIs) over UI actions. This framework also facilitates the creation and expansion of APIs through automated exploration of applications. Our experiments on Microsoft Word demonstrate that AXIS reduces task completion time by 65%-70% and cognitive workload by 38%-53%, while maintaining accuracy of 97%-98% compared to humans. Our work contributes to a new human-agent-computer interaction (HACI) framework and explores a fresh UI design principle for application providers to turn applications into agents in the era of LLMs, paving the way towards an agent-centric operating system (Agent OS).

cs.AI↗

arXiv2025

WorldPM: Scaling Human Preference Modeling

Motivated by scaling laws in language modeling that demonstrate how test loss scales as a power law with model and dataset sizes, we find that similar laws exist in preference modeling. We propose World Preference Modeling$ (WorldPM) to emphasize this scaling potential, where World Preference embodies a unified representation of human preferences. In this paper, we collect preference data from public forums covering diverse user communities, and conduct extensive training using 15M-scale data across models ranging from 1.5B to 72B parameters. We observe distinct patterns across different evaluation metrics: (1) Adversarial metrics (ability to identify deceptive features) consistently scale up with increased training data and base model size; (2) Objective metrics (objective knowledge with well-defined answers) show emergent behavior in larger language models, highlighting WorldPM's scalability potential; (3) Subjective metrics (subjective preferences from a limited number of humans or AI) do not demonstrate scaling trends. Further experiments validate the effectiveness of WorldPM as a foundation for preference fine-tuning. Through evaluations on 7 benchmarks with 20 subtasks, we find that WorldPM broadly improves the generalization performance across human preference datasets of varying sizes (7K, 100K and 800K samples), with performance gains exceeding 5% on many key subtasks. Integrating WorldPM into our internal RLHF pipeline, we observe significant improvements on both in-house and public evaluation sets, with notable gains of 4% to 8% in our in-house evaluations.

cs.CL↗

arXiv2025

High-efficient long-distance device-independent quantum secret sharing based on single-photon sources

Device-independent quantum secret sharing (DI QSS) relaxes security assumptions of experimental devices to provide the highest security level for QSS. Previous DI QSS protocols require to generate multi-partite entangled states and then implement the long-distance entanglement distribution. The extremely low generation probability of current multi-partite entanglement source and photon transmission loss largely limit DI QSS's practical key generation rate and secure photon transmission distance. In the paper, we propose a high-efficient long-distance DI QSS protocol based on practical nearly on-demand single-photon sources (SPSs), which adopts the single photons and heralded architecture to construct long-distance multi-partite entanglement channels. Our SPS DI QSS protocol can automatically eliminate the negative influence from photon transmission loss and largely increase the secure photon transmission distance. Moreover, we adopt active improvement strategies such as random key generation basis, postselection, and advanced postselection to further increase the practical key generation rate and lower the experimental requirement. Comparing with previous DI QSS protocols based on entanglement sources, our SPS DI QSS protocol can increase the practical communication efficiency by six orders of magnitude, and increase the secure photon transmission distance by at least 151 times. This SPS DI QSS protocol is feasible with the current experimental technologies. Our protocol makes it possible to realize high-efficient long-distance DI network in the near future.

quant-ph↗