arXiv subjects

Huanhuan Ma

Publications and source records attributed to Huanhuan Ma.

14 recordsLinked to original sources

arXiv2026

MemAudit: Auditing Long-Term Agent Memory via Hidden User-State Recovery

Long-term memory promises LLM agents that grow more capable across sessions, maintaining an accurate, evolving understanding of the user that interaction forms. In practice, however, this memory is evaluated mostly through downstream behavior, such as later answers, personalization quality, or task success, which tests that understanding only indirectly and leaves the memory artifact itself largely unaudited. We argue that long-term memory should instead be evaluated as an auditable post-interaction artifact: after ordinary assistance, what structured user state can be reconstructed from the memory the agent leaves behind? We instantiate this view in MEMPROBE, a benchmark in which a memory-equipped agent assists simulated users, each carrying a hidden, taxonomy-anchored user-state bank, across a trajectory of leak-controlled tasks, after which that bank is reconstructed from the agent's resulting memory under both full-store and top-k access. Built on synthetic ground truth for efficient, scalable measurement, MEMPROBE spans 50 simulated users with 31 hidden dimensions each (1,550 recovery targets) and tests 5 representative memory systems. Testing state-of-the-art memory agents, we find that successful assistance and recoverable memory behave as distinct capabilities. Task completion nearly saturates, even for a memoryless baseline, while category-balanced recovery stays moderate (about 0.6) and drops further under top-k retrieval. MEMPROBE is the first benchmark to study memory recovery directly, reconstructing the user state a system retains and scoring it against ground truth. We see recovery as a concrete objective for future memory agents to optimize, and MEMPROBE as a step toward an environment where agents are trained to remember their users, growing more faithful the longer they know them.

cs.CL↗

arXiv2026

Sycophancy Suppression Can Impair Rational Updating: Anti-Sycophancy Should Preserve the Ability to Update

Large language models often exhibit sycophancy, revising their answers to align with users when users push back. Such answer flips, however, can arise from different causes. One possibility is that the model simply aligns with the user's feedback in order to satisfy them. Another is that the feedback genuinely contains useful evidence, prompting the model to update its answer in a rational way. We distinguish them as Unsupported-Yielding and Rational-Updating. Prior work focuses primarily on suppressing Unsupported-Yielding, while overlooking its effect on Rational-Updating. We address this gap with a two-turn evaluation framework that measures the two behaviors separately. Across representative training-time and inference-time interventions, we find that anti-sycophancy methods often encounter a trade-off in which reducing Unsupported-Yielding can sacrifice Rational-Updating, and vice versa, even when the two objectives are optimized jointly. Mechanistic analysis suggests that the two behaviors share an internal substrate: the MLP neurons and attention heads driving them overlap substantially, and their associated steering directions are positively aligned. We further conduct a preliminary orthogonalized steering exploration, which yields modest, backbone-dependent selectivity gains. Overall, our results suggest that anti-sycophancy should be treated not as a simple suppression problem, but as a selectivity problem, where effective interventions should preserve Rational-Updating while reducing Unsupported-Yielding.

cs.CL↗

arXiv2026

Beyond BFI: The CSI for Enhanced Reliability and Validity in Evaluating LLM Personality Traits

As large language models (LLMs) increasingly function as human-like assistants exhibiting human-like personality traits, understanding their behavioral characteristics becomes essential for responsible AI development. However, existing evaluation efforts, which often adapt human psychological assessments such as the Big Five Inventory (BFI), face two significant limitations. First, these approaches often lack reliability, as minor prompt variations can lead to inconsistent test results. Second, the theoretical foundations of these tools, rooted in human studies, are misaligned with the computational nature of LLMs, thereby limiting their validity in predicting real-world model behavior. To address these limitations, we introduce the Core Sentiment Inventory (CSI), a novel personality trait evaluation instrument designed from the ground up and specifically tailored to the unique characteristics of LLMs. CSI covers both English and Chinese, that implicitly evaluates models' personality traits, providing insightful psychological portraits of LLMs. Extensive experiments demonstrate that: (1) CSI effectively captures nuanced behavioral patterns, revealing significant behavioral variations in LLMs across different languages and contexts; (2) Compared to current evaluation tools, CSI significantly improves reliability, yielding more consistent and robust results; and (3) The correlation between CSI scores and LLMs' real-world outputs exceeds 0.85, demonstrating its strong validity in predicting LLM behavior.

cs.CL↗

arXiv2026

A Deployment Audit of Release-Side Risk in Conformal Triage under Prevalence Shift

Conformal triage converts predictive scores into deployment actions that either release a case, flag it for urgent attention, or defer it to human review. Under an observed change in target-event prevalence, however, marginal coverage and human-review rate can miss whether patients who experience the target event are released without review. To address this gap, we introduce a leakage-aware deployment audit for release-side conformal triage. It first assigns target subjects to three non-overlapping roles: prevalence correction, conformal calibration, and held-out release-side evaluation. This separation then lets the audit evaluate release directly: how many event-positive patients are cleared without review, whether the pilot has enough event labels for calibration, and how the release-review trade-off shifts. Applying this audit to a retrospective non-small-cell lung cancer (NSCLC) target cohort shows why lower review can be misleading: after prevalence correction, the pooled conformal branch lowers review by releasing more patients, some of whom are event-positive. Within the audit, the classwise branch acts as a scarcity diagnostic: the pilot has too few event labels to support a low-review release rule.

cs.LG↗

arXiv2025

Infinity-MM: Scaling Multimodal Performance with Large-Scale and High-Quality Instruction Data

Recently, Vision-Language Models (VLMs) have achieved remarkable progress in multimodal tasks, and multimodal instruction data serves as the foundation for enhancing VLM capabilities. Despite the availability of several open-source multimodal datasets, limitations in the scale and quality of open-source instruction data hinder the performance of VLMs trained on these datasets, leading to a significant gap compared to models trained on closed-source data. To address this challenge, we introduce Infinity-MM, a large-scale multimodal instruction dataset. We collected the available multimodal instruction datasets and performed unified preprocessing, resulting in a dataset with over 40 million samples that ensures diversity and accuracy. Furthermore, to enable large-scale expansion of instruction data and support the continuous acquisition of high-quality data, we propose a synthetic instruction generation method based on a tagging system and open-source VLMs. By establishing correspondences between different types of images and associated instruction types, this method can provide essential guidance during data synthesis. Leveraging this high-quality data, we have trained a 2-billion-parameter Vision-Language Model, Aquila-VL-2B, which achieves state-of-the-art (SOTA) performance among models of similar scale. The data is available at: https://huggingface.co/datasets/BAAI/Infinity-MM.

cs.CL↗

arXiv2024

Training Data for Large Language Model

In 2022, with the release of ChatGPT, large-scale language models gained widespread attention. ChatGPT not only surpassed previous models in terms of parameters and the scale of its pretraining corpus but also achieved revolutionary performance improvements through fine-tuning on a vast amount of high-quality, human-annotated data. This progress has led enterprises and research institutions to recognize that building smarter and more powerful models relies on rich and high-quality datasets. Consequently, the construction and optimization of datasets have become a critical focus in the field of artificial intelligence. This paper summarizes the current state of pretraining and fine-tuning data for training large-scale language models, covering aspects such as data scale, collection methods, data types and characteristics, processing workflows, and provides an overview of available open-source datasets.

cs.AI↗

arXiv2024

Does Knowledge Localization Hold True? Surprising Differences Between Entity and Relation Perspectives in Language Models

Large language models encapsulate knowledge and have demonstrated superior performance on various natural language processing tasks. Recent studies have localized this knowledge to specific model parameters, such as the MLP weights in intermediate layers. This study investigates the differences between entity and relational knowledge through knowledge editing. Our findings reveal that entity and relational knowledge cannot be directly transferred or mapped to each other. This result is unexpected, as logically, modifying the entity or the relation within the same knowledge triplet should yield equivalent outcomes. To further elucidate the differences between entity and relational knowledge, we employ causal analysis to investigate how relational knowledge is stored in pre-trained models. Contrary to prior research suggesting that knowledge is stored in MLP weights, our experiments demonstrate that relational knowledge is also significantly encoded in attention modules. This insight highlights the multifaceted nature of knowledge storage in language models, underscoring the complexity of manipulating specific types of knowledge within these models.

cs.CL↗

arXiv2024

Navigating the Noisy Crowd: Finding Key Information for Claim Verification

Claim verification is a task that involves assessing the truthfulness of a given claim based on multiple evidence pieces. Using large language models (LLMs) for claim verification is a promising way. However, simply feeding all the evidence pieces to an LLM and asking if the claim is factual does not yield good results. The challenge lies in the noisy nature of both the evidence and the claim: evidence passages typically contain irrelevant information, with the key facts hidden within the context, while claims often convey multiple aspects simultaneously. To navigate this "noisy crowd" of information, we propose EACon (Evidence Abstraction and Claim Deconstruction), a framework designed to find key information within evidence and verify each aspect of a claim separately. EACon first finds keywords from the claim and employs fuzzy matching to select relevant keywords for each raw evidence piece. These keywords serve as a guide to extract and summarize critical information into abstracted evidence. Subsequently, EACon deconstructs the original claim into subclaims, which are then verified against both abstracted and raw evidence individually. We evaluate EACon using two open-source LLMs on two challenging datasets. Results demonstrate that EACon consistently and substantially improve LLMs' performance in claim verification.

cs.CL↗

arXiv2024

EX-FEVER: A Dataset for Multi-hop Explainable Fact Verification

Fact verification aims to automatically probe the veracity of a claim based on several pieces of evidence. Existing works are always engaging in accuracy improvement, let alone explainability, a critical capability of fact verification systems. Constructing an explainable fact verification system in a complex multi-hop scenario is consistently impeded by the absence of a relevant, high-quality dataset. Previous datasets either suffer from excessive simplification or fail to incorporate essential considerations for explainability. To address this, we present EXFEVER, a pioneering dataset for multi-hop explainable fact verification. With over 60,000 claims involving 2-hop and 3-hop reasoning, each is created by summarizing and modifying information from hyperlinked Wikipedia documents. Each instance is accompanied by a veracity label and an explanation that outlines the reasoning path supporting the veracity classification. Additionally, we demonstrate a novel baseline system on our EX-FEVER dataset, showcasing document retrieval, explanation generation, and claim verification, and validate the significance of our dataset. Furthermore, we highlight the potential of utilizing Large Language Models in the fact verification task. We hope our dataset could make a significant contribution by providing ample opportunities to explore the integration of natural language explanations in the domain of fact verification.

cs.AI↗

arXiv2024

A fast low-rank inversion algorithm of dielectric matrix in GW approximation

The dielectric response function and its inverse are crucial physical quantities in materials science. We propose an accurate and efficient strategy to invert the dielectric function matrix. The GW approximation, a powerful approach to accurately describe many-body excited states, is taken as an application to demonstrate accuracy and efficiency. We incorporate the interpolative separable density fitting (ISDF) algorithm with Sherman--Morrison--Woodbury (SMW) formula to accelerate the inversion process by exploiting low-rank properties of dielectric function in plane-wave GW calculations. Our ISDF--SMW strategy produces accurate quasiparticle energies with $O(N_{\mathrm{r}}N_{\mathrm{e}}^2)$ computational cost $(N_{\mathrm{e}}$ is the number of electrons and $N_{\mathrm{r}}=100$--$1000N_{\mathrm{e}}$ is the number of grid points) with negligible small error of $0.03$ eV for both complex molecules and solids. This new strategy for inverting the dielectric matrix can be $50\times$ faster than the current state-of-the-art implementation in BerkeleyGW, resulting in two orders of magnitude speedup for total GW calculations.

math.NA↗

arXiv2024

Interpretable Multimodal Out-of-context Detection with Soft Logic Regularization

The rapid spread of information through mobile devices and media has led to the widespread of false or deceptive news, causing significant concerns in society. Among different types of misinformation, image repurposing, also known as out-of-context misinformation, remains highly prevalent and effective. However, current approaches for detecting out-of-context misinformation often lack interpretability and offer limited explanations. In this study, we propose a logic regularization approach for out-of-context detection called LOGRAN (LOGic Regularization for out-of-context ANalysis). The primary objective of LOGRAN is to decompose the out-of-context detection at the phrase level. By employing latent variables for phrase-level predictions, the final prediction of the image-caption pair can be aggregated using logical rules. The latent variables also provide an explanation for how the final result is derived, making this fine-grained detection method inherently explanatory. We evaluate the performance of LOGRAN on the NewsCLIPpings dataset, showcasing competitive overall results. Visualized examples also reveal faithful phrase-level predictions of out-of-context images, accompanied by explanations. This highlights the effectiveness of our approach in addressing out-of-context detection and enhancing interpretability.

cs.CV↗

arXiv2023

Assessing Knowledge Editing in Language Models via Relation Perspective

Knowledge Editing (KE) for modifying factual knowledge in Large Language Models (LLMs) has been receiving increasing attention. However, existing knowledge editing methods are entity-centric, and it is unclear whether this approach is suitable for a relation-centric perspective. To address this gap, this paper constructs a new benchmark named RaKE, which focuses on Relation based Knowledge Editing. In this paper, we establish a suite of innovative metrics for evaluation and conduct comprehensive experiments involving various knowledge editing baselines. We notice that existing knowledge editing methods exhibit the potential difficulty in their ability to edit relations. Therefore, we further explore the role of relations in factual triplets within the transformer. Our research results confirm that knowledge related to relations is not only stored in the FFN network but also in the attention layers. This provides experimental support for future relation-based knowledge editing methods.

cs.CL↗

arXiv2023

MenatQA: A New Dataset for Testing the Temporal Comprehension and Reasoning Abilities of Large Language Models

Large language models (LLMs) have shown nearly saturated performance on many natural language processing (NLP) tasks. As a result, it is natural for people to believe that LLMs have also mastered abilities such as time understanding and reasoning. However, research on the temporal sensitivity of LLMs has been insufficiently emphasized. To fill this gap, this paper constructs Multiple Sensitive Factors Time QA (MenatQA), which encompasses three temporal factors (scope factor, order factor, counterfactual factor) with total 2,853 samples for evaluating the time comprehension and reasoning abilities of LLMs. This paper tests current mainstream LLMs with different parameter sizes, ranging from billions to hundreds of billions. The results show most LLMs fall behind smaller temporal reasoning models with different degree on these factors. In specific, LLMs show a significant vulnerability to temporal biases and depend heavily on the temporal information provided in questions. Furthermore, this paper undertakes a preliminary investigation into potential improvement strategies by devising specific prompts and leveraging external tools. These approaches serve as valuable baselines or references for future research endeavors.

cs.CL↗

arXiv2022

KSSOLV 2.0: An efficient MATLAB toolbox for solving the Kohn-Sham equations with plane-wave basis set

KSSOLV (Kohn-Sham Solver) is a MATLAB toolbox for performing Kohn-Sham density functional theory (DFT) calculations with a plane-wave basis set. KSSOLV 2.0 preserves the design features of the original KSSOLV software to allow users and developers to easily set up a problem and perform ground-state calculations as well as to prototype and test new algorithms. Furthermore, it includes new functionalities such as new iterative diagonalization algorithms, k-point sampling for electron band structures, geometry optimization and advanced algorithms for performing DFT calculations with local, semi-local, and hybrid exchange-correlation functionals. It can be used to study the electronic structures of both molecules and solids. We describe these new capabilities in this work through a few use cases. We also demonstrate the numerical accuracy and computational efficiency of KSSOLV on a variety of examples.

physics.comp-ph↗