Searcharxiv⌕ Search

arXiv subjects

Ou Wu

Publications and source records attributed to Ou Wu.

At least 19 recordsLinked to original sources
arXiv2026

Data Optimization for LLMs: A Survey

Large language models (LLMs) have attracted great attention in both academic and industrial communities in recent years. Numerous survey studies have covered various aspects of LLMs, such as learning strategies, applications, alignment, explainability, and evaluation. However, little headway has been made in reviewing data-related aspects for LLMs, despite data being recognized as a key factor contributing to the emergence of the LLMs era. A wide range of data optimization algorithms and techniques are employed in LLMs' training and application processes. These techniques often originate from diverse deep learning domains, with their theoretical inspirations and heuristic motivations appearing unrelated to each other. This study aims to develop a comprehensive framework for LLMs' data optimization techniques to enhance data utilization and efficiency. We first discuss data challenges in LLMs' training and inference. Second, we summarize data perception dimensions for evaluating LLMs' data properties. These dimensions primarily include quality, difficulty, diversity, and redundancy, laying a foundation for subsequent optimization. Third, existing optimization techniques are reviewed and categorized into four groups: data selection, augmentation, reweighting, and others. Fourth, we analyze interconnections between these categories. Finally, we identify key challenges and future directions for LLMs' data optimization compared with conventional deep learning approaches.

cs.AI↗
arXiv2026

MetaKE: Meta-Learning for Knowledge Editing Toward a Better Accuracy-Editability Trade-off

Existing locate-then-edit Knowledge Editing (KE) methods typically decompose editing into two stages: upstream target representation optimization and downstream constrained parameter optimization. The optimization across the two stages is disconnected: upstream applies uniform regularization without observing downstream realization of the planned residual, hindering a refined accuracy-editability trade-off. Since this realization is request-specific and depends on downstream constraints, uniform regularization can over-shrink high-association requests, causing insufficient editing, while it can under-regularize low-association requests, producing over-large planned residuals that reduce downstream editability. To bridge this disconnect, we propose MetaKE (Meta-learning for Knowledge Editing), a new framework that unifies upstream and downstream stages into a bi-level optimization problem. The inner level optimizes parameter updates for the target representation, while the outer level optimizes representation using feedback from downstream constraints, achieving a better semantic accuracy-editability trade-off. To avoid costly multi-layer backpropagation, we introduce a Structural Gradient Proxy to approximate and propagate this feedback. Extensive experiments show that MetaKE outperforms strong baselines, offering a new perspective on KE.

cs.CL↗
arXiv2026

DRIVE: Fine-tuning via Data Contribution- and Diversity-aware Weighting with Prior Regularization

Single-stage fine-tuning is pervasive, yet catastrophic forgetting persists when the source pre-training data is inaccessible. Data weighting affords a natural lever to modulate sample influence. Yet existing weighting methods lack a principled mechanism to explicitly arbitrate knowledge retention against new-task adaptation, and provide no unified differentiable objective to accommodate auxiliary signals such as sample diversity and sample characteristics. This is because current schemes privilege one goal over the other, and no explicit gradient-compatible encoding exists for such signals. To address this, we propose DRIVE, a principled and efficient weight assignment framework. It derives per-sample weights via a unified differentiable objective where three coupled signals co-evolve: Shapley additivity splits contribution into anti-forgetting and adaptation components, which are reshaped by a covariance-based diversity term and stabilized by a sample-characteristic prior. Far from a mere combination, this yields an end-to-end gradient-based scheme balancing stability against plasticity. Theoretically, we prove that DRIVE enjoys superior stability-plasticity guarantees over existing weighting schemes and enhances parameter estimation robustness. Empirically, it attains superior overall performance, faster convergence, and stronger robustness. Code at https://github.com/luciabcd/DRIVE.

cs.LG↗
arXiv2026

One Algorithm, Two Goals: Dual Scoring for Parameter and Data Selection in LLM Fine-Tuning

In Large Language Model (LLM) fine-tuning, parameter and data selection are common strategies for reducing fine-tuning cost, yet they are typically driven by separate scoring mechanisms. When a parameter mask and data subset jointly determine restricted fine-tuning, this separation incurs redundant overhead and makes coordinated selection difficult. We cast parameter and data selection as two bilevel selection problems under a common validation objective and derive a shared local response-surrogate scoring rule. Under first- and second-order validation-improvement approximations, parameter importance and data utility emerge as column-wise and row-wise aggregations of a single gradient interaction matrix, yielding a closed-form row-column correspondence for co-extracting both signals. Building on this structure, we propose DualSFT (Dual-Selection Fine-Tuning), a one-shot dual-scoring algorithm that produces a parameter mask and data subset from shared gradient statistics. On 3B-9B LLMs, single-axis DualSFT variants strengthen target-task performance and stability-plasticity trade-offs within their comparison groups, while full DualSFT yields a more favorable joint-constrained trade-off than sequential hybrid baselines under matched budgets.

cs.LG↗
arXiv2026

NADS: Navigator-Guided Data Selection for Mitigating Catastrophic Forgetting in Fine-Tuning

Catastrophic forgetting is a central obstacle in single-stage fine-tuning of large language models (LLMs) on a new task, especially when the original pretraining data are unavailable. Synthetic data offers a practical mitigation path, but candidate pools are often noisy and redundant, making effective selection essential. Existing methods often rely on indirect proxy signals whose alignment with actual training benefit can be limited, while utility--diversity selection over large candidate pools remains computationally demanding. To address these challenges, we propose NAvigator-guided Data Selection (NADS), a fine-tuning framework for LLMs. NADS first follows the target fine-tuning recipe without preservation regularization to obtain a navigator model, whose deviation from the pretrained model exposes recipe-induced capability drift. It then uses the predictive divergence between the navigator and the pretrained model to score forgetting-aware utility for each synthetic candidate, and constructs a constraint set through an efficient utility--diversity selection strategy. Finally, NADS fine-tunes on the new-task data while distilling pretrained behavior on the selected constraints to preserve general capabilities. Experiments show that NADS delivers a stronger balance between new-task performance and general capability preservation, while reducing utility--diversity selection overhead. Code at: https://github.com/zjh-vinky/NADS.

cs.LG↗
arXiv2026

Revisiting Ripple Effects in Knowledge Editing through Pressure-Aware Joint Neighborhood Optimization

Single-edit updates in large language models can trigger ripple effects across local knowledge neighborhoods: desirable propagation to related facts and unintended perturbation of preserved ones. Existing methods address these two effects separately, without explicitly modeling their coupling. We challenge this separation through an analysis of ripple responses across typical baselines, identifying two coupled design pressures: editable-side coordination and preserved-side leakage. We propose Joint Neighborhood Optimization (JNO), a new knowledge-editing framework to formalize and jointly address both pressures at the target-planning stage. JNO instantiates this principle through Pressure-Aware Coordination (PAC), which jointly optimizes neighborhood target representations under coupled constraints, and a semantic pre-execution gate that rejects high-risk target plans before parameter execution. Experiments on RippleEdits show JNO improves propagation and preservation metrics by at least 7.0% while preserving cross-backbone editing stability.

cs.AI↗
arXiv2026

Token Utility Is Selection-Conditioned: Coupled Selection of Prompt Context and Response Supervision for Instruction Tuning

Token selection for instruction tuning is coupled: supervision utility depends on retained context, and vice versa. Existing methods largely optimize prompt compression or response supervision separately. Combining selectors can value supervision using removed context, or context using excluded supervision, creating selection-state mismatch. BRIDGE (Budgeted Response-Prompt Interaction via Directional Gradient-guided Efficient Token Selection) models selection-conditioned token utility through a shared validation-directed surrogate objective. BRIDGE routes response utility to supporting prompt tokens through attention, then alternates budgeted selection to revalue each side under the other's retained subset. Updates reuse precomputed interactions without repeated model scoring, while structure-aware projection promotes coherent supervision spans. Across three model families and three task domains, BRIDGE performs best overall among evaluated selection methods. Its MATH advantage over independent selection grows with compression, reaching 1.66 percentage points on Llama-3.2-3B.

cs.LG↗
arXiv2026

One Distribution, Two Channels: Coverage-Calibrated KV Retention and Compensation

Long-context decoding is increasingly constrained by key--value (KV) cache memory and bandwidth. Existing fixed-budget compression methods typically separate retention from compensation, while a retention ranking specifies neither discarded attention mass nor the direction of induced output error. We start from an exact factorization: eviction error equals evicted attention mass times the directional gap between the evicted centroid and retained output, highlighting the importance of set-level coverage in retention and mass-preserving memory writing. We introduce CORE COverage Calibration and Evicted-Mass REdistribution for KV Cache, which distills an offline allocation combining query utility and log-determinant coverage into a lightweight cache-aware indexer. At inference, one calibrated distribution drives both channels: its Top-$B$ ordering retains complementary KV states, while its excluded allocation mass and conditional weights parameterize latent-memory writes without a separate write-weight predictor or online log-determinant evaluation. Our analysis provides a four-term pre-compensation error certificate, characterizes non-additive coverage interactions, and establishes mass-independent write stability with hierarchical bounds through recurrent and query-adaptive normalization. Across three backbones, CORE exceeds the strongest RULER baseline by up to 3.78 points at 90% compression; LongBench and repeated-eviction evaluations further demonstrate strong effectiveness and decoding efficiency.

cs.AR↗
arXiv2026

HiLoRe: What to Store, Compress, or Recompute for Efficient GRPO Training

Group-relative policy optimization (GRPO) makes learner-side activations a major memory-computation bottleneck: gradient checkpointing reduces activation memory through recomputation, but fixed schedules can leave roughly 18 GB unused on a 48-GB GPU despite substantial recomputation overhead. Existing activation-management methods set state fidelity from execution cost, tensor properties, or generic compression sensitivity, without explicitly incorporating GRPO's analytic update structure into state-fidelity allocation. We formalize this dependence as policy-update exposure, linking the current GRPO loss coefficients to state-level approximation sensitivity. These coefficients are available before backward without an additional backward pass. We introduce HiLoRe, which allocates graph-attributed recovery units among high-precision storage, low-precision compression, and deterministic recomputation using measured recovery utility and update-conditioned approximation risk. It combines high-precision storage and deterministic recomputation with low-precision recovery under a calibrated risk budget. Across five model-task settings with 2K responses and memory < 1.10 times GC's per-GPU actor-update peak, HiLoRe's actor-update throughput gains reach 13.5% over GC and 7.9% over the fastest evaluated baseline, with paired mean downstream-score differences below 0.6 percentage points.

cs.LG↗
arXiv2026

AlphaToken: Decoupling Adaptation and Stability for Path-Aware Response Token Valuation in LLM Post-Training

Token selection is pivotal for effective LLM post-training. However, existing methods mostly rely on local heuristics and rarely formulate token selection as a principled valuation of individual response tokens. We introduce $\textbf{AlphaToken}$, a response token valuation framework that decouples valuation into $\textbf{adaptation}$ (promoting target-task learning) and $\textbf{stability}$ (preserving pre-trained capabilities), and makes each objective $\textbf{path-aware}$ by combining the direct-path signal from local token gradients with the downstream causal-path signal in autoregressive generation. Since retention data are typically unavailable, AlphaToken approximates stability via a $\textbf{Fisher-drift proxy}$ anchored at the pre-trained reference model. For efficient computation, we extend Ghost Dot-Product to token-level valuation. AlphaToken masks low-value response tokens during fine-tuning and preference optimization, concentrating training signals on more valuable positions. Experiments show that AlphaToken improves post-training performance and mitigates catastrophic forgetting.

cs.CL↗
arXiv2026

Two to Tango: Coupled Task-Reference Selection for Safe LLM Fine-tuning

Fine-tuning safety aligned large language models (LLMs) on downstream data improves adaptation but may erode learned safety behavior. Existing methods use fixed safety examples, global constraints, or one-sided task filtering. Our diagnostics show task updates expose different safety constraints, motivating joint selection of relevant references and compatible task samples. We propose DualSelect, a coupled framework for task and reference selection that refreshes task conditioned safety references before filtering whole task samples compatible with the induced reference direction. Under a minimax view, DualSelect selects safety references with high preservation loss and task conflict, together with compatible task samples, through entropy-regularized scoring surrogates, lazy reference refresh, and gradient correction. On 1B-8B LLMs, DualSelect preserves safety without losing task utility; using the REDORCA judge, it improves Safety Avg. over the strongest baseline by at least 5.10 points and remains highest in Safety Avg. across judges with moderate overhead. This view extends to retention focused continual learning.

cs.LG↗
arXiv2026

Computational Challenges in Token Economics: Bridging Economic Theory and AI System Design

Token economics has emerged as a useful lens for understanding resource allocation, value creation, and pricing in large language model systems. While recent work has increasingly treated tokens as economic primitives, there remains a substantial gap between high-level economic theory and the computational realities of modern AI infrastructure. This paper identifies and analyzes the key computational challenges that arise when token-economic principles are implemented in real-time inference systems. We argue that computational feasibility is not merely one dimension of token economics, but its governing constraint: these challenges are driven by fundamental tensions among fine-grained valuation, low-latency execution, and allocation optimality under uncertainty. To structure this problem space, we introduce the notion of \textbf{Computational Token Economics} and propose the \textbf{Token Economics Trilemma} -- a conditional no-free-lunch principle that captures the inherent trade-offs among granularity, real-time performance, and optimality. We further categorize the main technical challenges into three areas: real-time value accounting, constrained resource allocation, and economic-aware system architecture. Rather than presenting a complete solution, this paper aims to define a research agenda for bridging token economics and AI system design, highlighting open problems at the intersection of computational economics, machine learning systems, and AI infrastructure.

cs.AI↗
arXiv2026

Towards a Data-Parameter Correspondence for LLMs: A Preliminary Discussion

Large language model optimization has historically bifurcated into isolated data-centric and model-centric paradigms: the former manipulates involved samples through selection, augmentation, or poisoning, while the latter tunes model weights via masking, quantization, or low-rank adaptation. This paper establishes a unified \emph{data-parameter correspondence} revealing these seemingly disparate operations as dual manifestations of the same geometric structure on the statistical manifold $\mathcal{M}$. Grounded in the Fisher-Rao metric $g_{ij}(θ)$ and Legendre duality between natural ($θ$) and expectation ($η$) parameters, we identify three fundamental correspondences spanning the model lifecycle: 1. Geometric correspondence: data pruning and parameter sparsification equivalently reduce manifold volume via dual coordinate constraints; 2. Low-rank correspondence: in-context learning (ICL) and LoRA adaptation explore identical subspaces on the Grassmannian $\mathcal{G}(r,d)$, with $k$-shot samples geometrically equivalent to rank-$r$ updates; 3. Security-privacy correspondence: adversarial attacks exhibit cooperative amplification between data poisoning and parameter backdoors, whereas protective mechanisms follow cascading attenuation where data compression multiplicatively enhances parameter privacy. Extending from training through post-training compression to inference, this framework provides mathematical formalization for cross-community methodology transfer, demonstrating that cooperative optimization integrating data and parameter modalities may outperform isolated approaches across efficiency, robustness, and privacy dimensions.

cs.LG↗
arXiv2025

Data Valuation by Fusing Global and Local Statistical Information

Data valuation has garnered increasing attention in recent years, given the critical role of high-quality data in various applications. Among diverse data valuation approaches, Shapley value-based methods are predominant due to their strong theoretical grounding. However, the exact computation of Shapley values is often computationally prohibitive, prompting the development of numerous approximation techniques. Despite notable advancements, existing methods generally neglect the incorporation of value distribution information and fail to account for dynamic data conditions, thereby compromising their performance and application potential. In this paper, we highlight the crucial role of both global and local statistical properties of value distributions in the context of data valuation for machine learning. First, we conduct a comprehensive analysis of these distributions across various simulated and real-world datasets, uncovering valuable insights and key patterns. Second, we propose an enhanced data valuation method that fuses the explored distribution characteristics into two regularization terms to refine Shapley value estimation. The proposed regularizers can be seamlessly incorporated into various existing data valuation methods. Third, we introduce a novel approach for dynamic data valuation that infers updated data values without recomputing Shapley values, thereby significantly improving computational efficiency. Extensive experiments have been conducted across a range of tasks, including Shapley value estimation, value-based data addition and removal, mislabeled data detection, and dynamic data valuation. The results showcase the consistent effectiveness and efficiency of our proposed methodologies, affirming the significant potential of global and local value distributions in data valuation.

cs.LG↗
arXiv2025

Implicit Counterfactual Data Augmentation for Robust Learning

Machine learning models are prone to capturing the spurious correlations between non-causal attributes and classes, with counterfactual data augmentation being a promising direction for breaking these spurious associations. However, generating counterfactual data explicitly poses a challenge, and incorporating augmented data into the training process decreases training efficiency. This study proposes an Implicit Counterfactual Data Augmentation (ICDA) method to remove spurious correlations and make stable predictions. Specifically, first, a novel sample-wise augmentation strategy is developed that generates semantically and counterfactually meaningful deep features with distinct augmentation strength for each sample. Second, we derive an easy-to-compute surrogate loss on the augmented feature set when the number of augmented samples becomes infinite. Third, two concrete schemes are proposed, including direct quantification and meta-learning, to derive the key parameters for the robust loss. In addition, ICDA is explained from a regularization perspective, revealing its capacity to improve intra-class compactness and augment margins at both class and sample levels. Extensive experiments have been conducted across various biased learning scenarios covering both image and text datasets, demonstrating that ICDA consistently enhances the generalization and robustness performance of popular networks.

cs.LG↗
arXiv2025

Data Poisoning in Deep Learning: A Survey

Deep learning has become a cornerstone of modern artificial intelligence, enabling transformative applications across a wide range of domains. As the core element of deep learning, the quality and security of training data critically influence model performance and reliability. However, during the training process, deep learning models face the significant threat of data poisoning, where attackers introduce maliciously manipulated training data to degrade model accuracy or lead to anomalous behavior. While existing surveys provide valuable insights into data poisoning, they generally adopt a broad perspective, encompassing both attacks and defenses, but lack a dedicated, in-depth analysis of poisoning attacks specifically in deep learning. In this survey, we bridge this gap by presenting a comprehensive and targeted review of data poisoning in deep learning. First, this survey categorizes data poisoning attacks across multiple perspectives, providing an in-depth analysis of their characteristics and underlying design princinples. Second, the discussion is extended to the emerging area of data poisoning in large language models(LLMs). Finally, we explore critical open challenges in the field and propose potential research directions to advance the field further. To support further exploration, an up-to-date repository of resources on data poisoning in deep learning is available at https://github.com/Pinlong-Zhao/Data-Poisoning.

cs.CR↗
arXiv2024

Is Data Valuation Learnable and Interpretable?

Measuring the value of individual samples is critical for many data-driven tasks, e.g., the training of a deep learning model. Recent literature witnesses the substantial efforts in developing data valuation methods. The primary data valuation methodology is based on the Shapley value from game theory, and various methods are proposed along this path. {Even though Shapley value-based valuation has solid theoretical basis, it is entirely an experiment-based approach and no valuation model has been constructed so far.} In addition, current data valuation methods ignore the interpretability of the output values, despite an interptable data valuation method is of great helpful for applications such as data pricing. This study aims to answer an important question: is data valuation learnable and interpretable? A learned valuation model have several desirable merits such as fixed number of parameters and knowledge reusability. An intrepretable data valuation model can explain why a sample is valuable or invaluable. To this end, two new data value modeling frameworks are proposed, in which a multi-layer perception~(MLP) and a new regression tree are utilized as specific base models for model training and interpretability, respectively. Extensive experiments are conducted on benchmark datasets. {The experimental results provide a positive answer for the question.} Our study opens up a new technical path for the assessing of data values. Large data valuation models can be built across many different data-driven tasks, which can promote the widespread application of data valuation.

cs.LG↗
arXiv2023

Data Optimization in Deep Learning: A Survey

Large-scale, high-quality data are considered an essential factor for the successful application of many deep learning techniques. Meanwhile, numerous real-world deep learning tasks still have to contend with the lack of sufficient amounts of high-quality data. Additionally, issues such as model robustness, fairness, and trustworthiness are also closely related to training data. Consequently, a huge number of studies in the existing literature have focused on the data aspect in deep learning tasks. Some typical data optimization techniques include data augmentation, logit perturbation, sample weighting, and data condensation. These techniques usually come from different deep learning divisions and their theoretical inspirations or heuristic motivations may seem unrelated to each other. This study aims to organize a wide range of existing data optimization methodologies for deep learning from the previous literature, and makes the effort to construct a comprehensive taxonomy for them. The constructed taxonomy considers the diversity of split dimensions, and deep sub-taxonomies are constructed for each dimension. On the basis of the taxonomy, connections among the extensive data optimization methods for deep learning are built in terms of four aspects. We probe into rendering several promising and interesting future directions. The constructed taxonomy and the revealed connections will enlighten the better understanding of existing methods and the design of novel data optimization techniques. Furthermore, our aspiration for this survey is to promote data optimization as an independent subdivision of deep learning. A curated, up-to-date list of resources related to data optimization in deep learning is available at \url{https://github.com/YaoRujing/Data-Optimization}.

cs.LG↗