arXiv subjects

Zifan Liu

Publications and source records attributed to Zifan Liu.

16 recordsLinked to original sources

arXiv2026

MAFIA: Query-Only Memory Attacks via Probing and Factual Injection against Audited LLM Agents

Memory-augmented LLM agents rely on rich context for long-horizon reasoning and acting, yet their memory modules expose a persistent attack surface for malicious records, making the study of memory poisoning threats imperative. However, existing query-only attacks often fail to remain effective in two realistic and prevalent settings: large-scale benign memory pools and active input auditing. Consequently, current approaches fall short when facing the dual challenges of high retrieval competitiveness and rigorous semantic checks. To overcome these limitations, we propose MAFIA, a query-only Memory Attack framework via probing and Factual Injection against Audit, tailored to this extended threat model. Specifically, MAFIA introduces: (1) a placement strategy that ensures retrieval-competitive injection via memory probing, budget allocation, and scheduling; and (2) a payload design that bypasses audits using compact factual cloaks, preserving malicious effects while maintaining high semantic similarity. Extensive evaluations reveal that MAFIA achieves up to a 90.7% attack success rate while suppressing audit detection from a peak of 83.3% to at most 7.4%, exposing critical vulnerabilities across agentic memory systems. Code will be made publicly available at https://github.com/JiamingChen1234/MAFIA.

cs.AI↗

arXiv2026

TACO: Task-Aware Column Description Generation Using LLMs

Generating accurate and informative column descriptions (e.g. "membership status of customers" for the column name "cust_mem") is essential for a wide range of downstream NLP tasks on tabular data, including NL2SQL, table question answering, and entity linking. This problem arises in enterprises, domain sciences, government data portals, and so on. Despite its importance, most real-world datasets suffer from missing or cryptic documentation, often due to abbreviated column names or domain-specific jargon. Existing approaches largely rely on single-prompt large language models (LLMs), which struggle with three key issues: (i) inconsistent or incorrect handling of abbreviations, (ii) hallucinated or incomplete descriptions, and (iii) redundancy or vagueness that hinders downstream performance. We present TACO, a task-aware framework for automatic column description generation using LLMs. TACO introduces a three-step pipeline: (1) abbreviation expansion, which standardizes column names; (2) description generation, which produces initial semantic descriptions enriched with synonyms and search-oriented keywords; and (3) description revision, which refines these outputs using simulated downstream tasks. In addition, we investigate human-in-the-loop extensions and release new evaluation datasets for entity linking and schema enrichment. Extensive experiments across public and proprietary datasets show that TACO consistently outperforms existing methods, improving downstream task performance by up to 32%.

cs.CL↗

arXiv2026

GEAR: Granularity-Adaptive Advantage Reweighting for LLM Agents via Self-Distillation

Reinforcement learning has become a widely used post-training approach for LLM agents, where training commonly relies on outcome-level rewards that provide only coarse supervision. While finer-grained credit assignment is promising for effective policy updates, obtaining reliable local credit and assigning it to the right parts of the long-horizon trajectory remains an open challenge. In this paper, we propose Granularity-adaptivE Advantage Reweighting (GEAR), an adaptive-granularity credit assignment framework that reshapes the trajectory-level GRPO advantage using token- and segment-level signals derived from self-distillation. GEAR compares an on-policy student with a ground-truth-conditioned teacher to obtain a reference-guided divergence signal for identifying adaptive segment boundaries and modulating local advantage weights. This divergence often spikes at the onset of a semantic deviation, while later tokens in the same autoregressive continuation may return to low divergence. GEAR therefore treats such spikes as anchors for adaptive credit regions: where the student remains aligned with the teacher, token-level resolution is preserved; where it departs, GEAR groups the corresponding continuation into an adaptive segment and uses the divergence at the departure point to modulate the segment' s advantage. Experiments across eight mathematical reasoning and agentic tool-use benchmarks with Qwen3 4B and 8B models show that GEAR consistently outperforms standard GRPO, self-distillation-only baselines, and token- or turn-level credit-assignment methods. The gains are especially strong on benchmarks with lower GRPO baseline accuracy, reaching up to around 20\% over GRPO, suggesting that the proposed adaptive reweighting scheme is especially useful in more challenging long-horizon settings.

cs.LG↗

arXiv2026

ISEE: Interactive Semantic Enrichment for Database Fields

LLM-based agents are increasingly being deployed for data-related tasks, including data sense-making, exploration, and retrieval. However, their performance heavily depends on the clarity and completeness of data semantics. In practice, many field descriptions remain ambiguous or incomplete, as much of the essential context (e.g., the meaning of a customized field) originates from users' domain knowledge and is rarely documented publicly. This gap restricts the agents' task performance in downstream tasks, such as entity-linking. To bridge this gap, we introduce a novel and comprehensive Interactive SEmantic Enrichment system (ISEE). Given a data field description, ISEE measures its quality through a scoring system, gathers domain knowledge, and collaboratively enriches the semantics with users. Through a user study, automated user simulation, quantitative evaluation, and case study, we demonstrate that ISEE significantly reduces cognitive load, improves description quality, and enhances downstream task performance.

cs.AI↗

arXiv2026

GAS: Enhancing Reward-Cost Balance of Generative Model-assisted Offline Safe RL

Offline Safe Reinforcement Learning (OSRL) aims to learn a policy to achieve high performance in sequential decision-making while satisfying constraints, using only pre-collected datasets. Recent works, inspired by the strong capabilities of Generative Models (GMs), reformulate decision-making in OSRL as a conditional generative process, where GMs generate desirable actions conditioned on predefined reward and cost values. However, GM-assisted methods face two major challenges in OSRL: (1) lacking the ability to "stitch" optimal transitions from suboptimal trajectories within the dataset, and (2) struggling to balance reward targets with cost targets, particularly when they are conflict. To address these issues, we propose Goal-Assisted Stitching (GAS), a novel algorithm designed to enhance stitching capabilities while effectively balancing reward maximization and constraint satisfaction. To enhance the stitching ability, GAS first augments and relabels the dataset at the transition level, enabling the construction of high-quality trajectories from suboptimal ones. GAS also introduces novel goal functions, which estimate the optimal achievable reward and cost goals from the dataset. These goal functions, trained using expectile regression on the relabeled and augmented dataset, allow GAS to accommodate a broader range of reward-cost return pairs and achieve a better tradeoff between reward maximization and constraint satisfaction compared to human-specified values. The estimated goals then guide policy training, ensuring robust performance under constrained settings. Furthermore, to improve training stability and efficiency, we reshape the dataset to achieve a more uniform reward-cost return distribution. Empirical results validate the effectiveness of GAS, demonstrating superior performance in balancing reward maximization and constraint satisfaction compared to existing methods.

cs.LG↗

arXiv2025

Experience-Evolving Multi-Turn Tool-Use Agent with Hybrid Episodic-Procedural Memory

As intents unfold and environments change, multi-turn agents face continuously shifting decision contexts. Although reusing past experience is intuitively appealing, existing approaches remain limited: full trajectories are often too context-specific to transfer, while tool-level reuse ignores the surrounding context and environment. In this paper, we introduce a hybrid episodic-procedural memory strategy (H-EPM) that enables experience-induced self-evolution of multi-turn tool-use policies by adaptively reusing partially overlapping successful experiences during both inference and training. Inspired by human episodic-procedural integration, we construct a tool graph from accumulated trajectories, where recurring tool-to-tool dependencies capture procedural routines and each edge is augmented with compact episodic summaries of relevant context. At inference time, the agent dynamically balances episodic recall for contextual reasoning with procedural execution for routine steps. Beyond inference, H-EPM introduces a memory-guided reinforcement learning paradigm that directly addresses a core challenge in multi-turn agent reinforcement learning, namely ineffective exploration over long trajectories. By biasing exploration toward historically successful tool transitions, H-EPM learns a stronger policy that generalizes at inference time without relying on domain-specific experience collection. Experiments show that H-EPM consistently delivers substantial inference-time gains over strong baselines across multi-turn tool-use benchmarks, reaching improvements of up to fifty percent. It also improves reinforcement learning policy performance, achieving gains of up to forty percent on out-of-distribution tasks.

cs.LG↗

arXiv2025

Structure-Aware Automatic Channel Pruning by Searching with Graph Embedding

Channel pruning is a powerful technique to reduce the computational overhead of deep neural networks, enabling efficient deployment on resource-constrained devices. However, existing pruning methods often rely on local heuristics or weight-based criteria that fail to capture global structural dependencies within the network, leading to suboptimal pruning decisions and degraded model performance. To address these limitations, we propose a novel structure-aware automatic channel pruning (SACP) framework that utilizes graph convolutional networks (GCNs) to model the network topology and learn the global importance of each channel. By encoding structural relationships within the network, our approach implements topology-aware pruning and this pruning is fully automated, reducing the need for human intervention. We restrict the pruning rate combinations to a specific space, where the number of combinations can be dynamically adjusted, and use a search-based approach to determine the optimal pruning rate combinations. Extensive experiments on benchmark datasets (CIFAR-10, ImageNet) with various models (ResNet, VGG16) demonstrate that SACP outperforms state-of-the-art pruning methods on compression efficiency and competitive on accuracy retention.

cs.AI↗

arXiv2025

ECLAIR: Enhanced Clarification for Interactive Responses

We present ECLAIR (Enhanced CLArification for Interactive Responses), a novel unified and end-to-end framework for interactive disambiguation in enterprise AI assistants. ECLAIR generates clarification questions for ambiguous user queries and resolves ambiguity based on the user's response.We introduce a generalized architecture capable of integrating ambiguity information from multiple downstream agents, enhancing context-awareness in resolving ambiguities and allowing enterprise specific definition of agents. We further define agents within our system that provide domain-specific grounding information. We conduct experiments comparing ECLAIR to few-shot prompting techniques and demonstrate ECLAIR's superior performance in clarification question generation and ambiguity resolution.

cs.AI↗

arXiv2025

ECLAIR: Enhanced Clarification for Interactive Responses in an Enterprise AI Assistant

Large language models (LLMs) have shown remarkable progress in understanding and generating natural language across various applications. However, they often struggle with resolving ambiguities in real-world, enterprise-level interactions, where context and domain-specific knowledge play a crucial role. In this demonstration, we introduce ECLAIR (Enhanced CLArification for Interactive Responses), a multi-agent framework for interactive disambiguation. ECLAIR enhances ambiguous user query clarification through an interactive process where custom agents are defined, ambiguity reasoning is conducted by the agents, clarification questions are generated, and user feedback is leveraged to refine the final response. When tested on real-world customer data, ECLAIR demonstrates significant improvements in clarification question generation compared to standard few-shot methods.

cs.CL↗

arXiv2025

Reinforcement Learning with Intrinsically Motivated Feedback Graph for Lost-sales Inventory Control

Reinforcement learning (RL) has proven to be well-performed and general-purpose in the inventory control (IC). However, further improvement of RL algorithms in the IC domain is impeded due to two limitations of online experience. First, online experience is expensive to acquire in real-world applications. With the low sample efficiency nature of RL algorithms, it would take extensive time to train the RL policy to convergence. Second, online experience may not reflect the true demand due to the lost sales phenomenon typical in IC, which makes the learning process more challenging. To address the above challenges, we propose a decision framework that combines reinforcement learning with feedback graph (RLFG) and intrinsically motivated exploration (IME) to boost sample efficiency. In particular, we first take advantage of the inherent properties of lost-sales IC problems and design the feedback graph (FG) specially for lost-sales IC problems to generate abundant side experiences aid RL updates. Then we conduct a rigorous theoretical analysis of how the designed FG reduces the sample complexity of RL methods. Based on the theoretical insights, we design an intrinsic reward to direct the RL agent to explore to the state-action space with more side experiences, further exploiting FG's power. Experimental results demonstrate that our method greatly improves the sample efficiency of applying RL in IC. Our code is available at https://anonymous.4open.science/r/RLIMFG4IC-811D/

cs.LG↗

arXiv2024

TSDS: Data Selection for Task-Specific Model Finetuning

Finetuning foundation models for specific tasks is an emerging paradigm in modern machine learning. The efficacy of task-specific finetuning largely depends on the selection of appropriate training data. We present TSDS (Task-Specific Data Selection), a framework to select data for task-specific model finetuning, guided by a small but representative set of examples from the target task. To do so, we formulate data selection for task-specific finetuning as an optimization problem with a distribution alignment loss based on optimal transport to capture the discrepancy between the selected data and the target distribution. In addition, we add a regularizer to encourage the diversity of the selected data and incorporate kernel density estimation into the regularizer to reduce the negative effects of near-duplicates among the candidate data. We connect our optimization problem to nearest neighbor search and design efficient algorithms to compute the optimal solution based on approximate nearest neighbor search techniques. We evaluate our method on data selection for both continued pretraining and instruction tuning of language models. We show that instruction tuning using data selected by our method with a 1% selection ratio often outperforms using the full dataset and beats the baseline selection methods by 1.5 points in F1 score on average.

cs.LG↗

arXiv2024

Individual Contributions as Intrinsic Exploration Scaffolds for Multi-agent Reinforcement Learning

In multi-agent reinforcement learning (MARL), effective exploration is critical, especially in sparse reward environments. Although introducing global intrinsic rewards can foster exploration in such settings, it often complicates credit assignment among agents. To address this difficulty, we propose Individual Contributions as intrinsic Exploration Scaffolds (ICES), a novel approach to motivate exploration by assessing each agent's contribution from a global view. In particular, ICES constructs exploration scaffolds with Bayesian surprise, leveraging global transition information during centralized training. These scaffolds, used only in training, help to guide individual agents towards actions that significantly impact the global latent state transitions. Additionally, ICES separates exploration policies from exploitation policies, enabling the former to utilize privileged global information during training. Extensive experiments on cooperative benchmark tasks with sparse rewards, including Google Research Football (GRF) and StarCraft Multi-agent Challenge (SMAC), demonstrate that ICES exhibits superior exploration capabilities compared with baselines. The code is publicly available at https://github.com/LXXXXR/ICES.

cs.LG↗

arXiv2023

Rapidash: Efficient Constraint Discovery via Rapid Verification

Denial Constraint (DC) is a well-established formalism that captures a wide range of integrity constraints commonly encountered, including candidate keys, functional dependencies, and ordering constraints, among others. Given their significance, there has been considerable research interest in achieving fast verification and discovery of exact DCs within the database community. Despite the significant advancements in the field, prior work exhibits notable limitations when confronted with large-scale datasets. The current state-of-the-art exact DC verification algorithm demonstrates a quadratic (worst-case) time complexity relative to the dataset's number of rows. In the context of DC discovery, existing methodologies rely on a two-step algorithm that commences with an expensive data structure-building phase, often requiring hours to complete even for datasets containing only a few million rows. Consequently, users are left without any insights into the DCs that hold on their dataset until this lengthy building phase concludes. In this paper, we introduce Rapidash, a comprehensive framework for DC verification and discovery. Our work makes a dual contribution. First, we establish a connection between orthogonal range search and DC verification. We introduce a novel exact DC verification algorithm that demonstrates near-linear time complexity, representing a theoretical improvement over prior work. Second, we propose an anytime DC discovery algorithm that leverages our novel verification algorithm to gradually provide DCs to users, eliminating the need for the time-intensive building phase observed in prior work. To validate the effectiveness of our algorithms, we conduct extensive evaluations on four large-scale production datasets. Our results reveal that our DC verification algorithm achieves up to 40 times faster performance compared to state-of-the-art approaches.

cs.DB↗

arXiv2022

AutoSlicer: Scalable Automated Data Slicing for ML Model Analysis

Automated slicing aims to identify subsets of evaluation data where a trained model performs anomalously. This is an important problem for machine learning pipelines in production since it plays a key role in model debugging and comparison, as well as the diagnosis of fairness issues. Scalability has become a critical requirement for any automated slicing system due to the large search space of possible slices and the growing scale of data. We present Autoslicer, a scalable system that searches for problematic slices through distributed metric computation and hypothesis testing. We develop an efficient strategy that reduces the search space through pruning and prioritization. In the experiments, we show that our search strategy finds most of the anomalous slices by inspecting a small portion of the search space.

cs.LG↗

arXiv2021

Picket: Guarding Against Corrupted Data in Tabular Data during Learning and Inference

Data corruption is an impediment to modern machine learning deployments. Corrupted data can severely bias the learned model and can also lead to invalid inferences. We present, Picket, a simple framework to safeguard against data corruptions during both training and deployment of machine learning models over tabular data. For the training stage, Picket identifies and removes corrupted data points from the training data to avoid obtaining a biased model. For the deployment stage, Picket flags, in an online manner, corrupted query points to a trained machine learning model that due to noise will result in incorrect predictions. To detect corrupted data, Picket uses a self-supervised deep learning model for mixed-type tabular data, which we call PicketNet. To minimize the burden of deployment, learning a PicketNet model does not require any human-labeled data. Picket is designed as a plugin that can increase the robustness of any machine learning pipeline. We evaluate Picket on a diverse array of real-world data considering different corruption models that include systematic and adversarial noise during both training and testing. We show that Picket consistently safeguards against corrupted data during both training and deployment of various models ranging from SVMs to neural networks, beating a diverse array of competing methods that span from data quality validation models to robust outlier-detection models.

cs.LG↗

arXiv2021

On Robust Mean Estimation under Coordinate-level Corruption

We study the problem of robust mean estimation and introduce a novel Hamming distance-based measure of distribution shift for coordinate-level corruptions. We show that this measure yields adversary models that capture more realistic corruptions than those used in prior works, and present an information-theoretic analysis of robust mean estimation in these settings. We show that for structured distributions, methods that leverage the structure yield information theoretically more accurate mean estimation. We also focus on practical algorithms for robust mean estimation and study when data cleaning-inspired approaches that first fix corruptions in the input data and then perform robust mean estimation can match the information theoretic bounds of our analysis. We finally demonstrate experimentally that this two-step approach outperforms structure-agnostic robust estimation and provides accurate mean estimation even for high-magnitude corruption.

cs.LG↗