arXiv subjects

Chen Xing

Publications and source records attributed to Chen Xing.

At least 19 recordsLinked to original sources

arXiv2026

Rethinking Text-Based Image Retrieval in Specific Domain

Driven by the rapid advancement of vision-language representation learning, Text-based Image Retrieval (TBIR) has made notable progress. However, existing benchmarks are predominantly constructed on an exclusive single-match assumption between query and images. While effective in general scenarios, this assumption fails to reflect practical system performance in specific domains (e.g., surveillance), where a single query often corresponds to multiple relevant candidate images. To address this limitation, we design a Domain-Specific Multi-Match Text-based Image Retrieval (DSMM-TBIR) data engine. Leveraging this engine, we construct Security Multi-Match TBIR (SecMM-TBIR), a benchmark comprising 50k surveillance images with 200 comprehensive queries. Furthermore, we observe that vanilla contrastive learning in specific domains suffers from severe false negatives, forcing the model to push apart semantically similar pairs and thus degrading retrieval performance. We propose the Semantic-Aware Fine-Tuning (SAFT) framework to address semantic compression in specific domains, which incorporates Semantic-Aware Soft-Label Supervision (SASS) and Intra-modal Structural Distillation (ISD) to establish a promising paradigm for domain-specific TBIR tasks. Experiments across diverse CLIP-like models demonstrate that SAFT yields an average mAP@20 gain of 7.8 points on SecMM-TBIR over standard image-text contrastive (ITC) fine-tuning, while also improving general-domain performance. The entire benchmark will be released to facilitate further research.

cs.CV↗

arXiv2026

Back Reaction of the Untwisting Solar Corona Scars Sunspots

The evolution of magnetic fields in the tenuous solar corona is predominantly governed by the motions of the underlying dense photosphere. Despite, coronal magnetic restructuring driven by magnetic reconnection between interacting coronal fields can sometimes react backwards to change photospheric magnetic fields. However, the mechanism of reactions remains undetermined. Here, we report the discovery of a back-reaction phenomenon: the untwisting of coronal loops that become twisted during reconnection in an eruption results in enhanced currents at the boundary of their footpoint away from the eruption, manifesting as the growth of a sunspot scar. It is revealed to arise from the Alfvenic reverse transfer of magnetic twist from the corona to the lower atmosphere, thanks to joint space observations and a magnetohydrodynamics simulation. These findings provide a viable and quantitative interpretation for the majority of puzzling photospheric changes associated with coronal mass ejections and/or flares and warn for unexpected magnetic field evolutions in sunspots and starspots.

astro-ph.SR↗

arXiv2026

Muse Spark Safety & Preparedness Report

Muse Spark is the latest large language model developed by Meta. In this report, we first present evaluations for catastrophic risk domains under Meta's Advanced AI Scaling Framework, along with the evidence that informed our launch decision. We then discuss additional considerations, such as Muse Spark's broader content safety and behavioral profile, that are relevant to overall safety but fall outside the catastrophic risk domains governed by the Framework. Our preparedness results covering Chemical and Biological, Cybersecurity, and Loss of Control risks assess Muse Spark's deployment within Meta AI as presenting acceptable levels of residual risks under our Advanced AI Scaling Framework. We conducted a broad set of evaluations targeting dual-use and high-risk capabilities across these catastrophic risk domains. Those evaluations identified elevated risks prior to mitigations, with Chemical and Biological capabilities assessed as likely reaching the "high risk" category under the Advanced AI Scaling Framework before safeguards were applied. We have implemented a multi-layered set of mitigations that address the identified risks, and Muse Spark demonstrates state-of-the-art refusal across a range of benchmarks related to hazardous workflows in chemistry and biology. We therefore release Muse Spark as the underlying model of Meta AI.

cs.CY↗

arXiv2025

Formation of a Magnetic Flux Rope Prior to the Eruption: Insight from a Radiative MHD Simulation of Active Region Emergence

Magnetic flux ropes (MFRs) are fundamental magnetic structures in solar eruptions, whose formation is generally attributed to (1) the emergence of subsurface flux tubes or (2) flux cancellation driven by photospheric horizontal flows and magnetic reconnection. Both mechanisms can operate simultaneously during active region evolution, making their relative contributions challenging to quantify. Here, we analyze the formation of a flux rope in a MURaM radiative magnetohydrodynamic (RMHD) simulation, which formed and evolved for approximately three hours before an M-class flare. The formation process is quantified by magnetic helicity flux, which drives the non-potential evolution of magnetic field, with its advection and shear terms on the photosphere corresponding to the emergence and photospheric horizontal flows, respectively. Examining the helicity injected into the flux rope through the photosphere, we find both terms increase significantly as the eruption approaches, with the shear term prevailing overall. Height-dependent analysis of helicity flux, together with magnetic field and velocity distributions, further reveals a gradual transition from the shear to the advection term with an increasing altitude, which is driven by magnetic reconnection above the photosphere. Our results provide quantitative evidence that flux cancellation governs flux rope formation, arising naturally from magnetic field reorganization during active region evolution: as flux emergence transports magnetic flux upward, photospheric shearing motions adjust magnetic field and inject helicity into solar atmosphere, and magnetic reconnection ultimately assembles the main body of flux ropes.

astro-ph.SR↗

arXiv2025

High-Resolution Modelling of Coronae and Winds in Solar-type Stars with Varying Rotation Rates I. X-ray Coronae

Stellar coronae are believed to be the main birthplace of various stellar magnetic activities. However, the structures and properties of stellar coronae remain poorly understood. Using the Space Weather Modelling Framework with the Alfv\'{e}n Wave Solar Model (SWMF-AWSoM) and dynamo-generated surface magnetic maps, here we model the coronae of four solar-type stars. By incorporating the Sun, our work covers a range of stars with the rotation varying from 1.0 to 23.3 $\Omega_\odot$ (periods of 25 to 1 days). Guided by observations, we scale the magnetic field strength with increasing rotation, covering a range between 6.0 G to 1200 G approximately. In our models, energy release associated with small-scale magnetic flux is a key source of coronal heating and is essential for reproducing realistic coronal structures. Our models capture dense (1$-$2 orders of magnitude higher than solar values) and ultra-hot ($\sim 10\,\mathrm{MK}$) coronae dominated by closed field structures. Using the CHIANTI atomic database, we also compute synthetic X-ray spectra and derive the corresponding X-ray luminosities $(L_X)$, which follow a scaling law to magnetic field $L_X \propto \langle|\mathbf{B}|\rangle^{1.75}$. Furthermore, the coronal X-ray emission is found to be rotationally modulated by the alternating presence of bright active regions and dark coronal holes. These results provide new insights into the extremely high-energy coronae of rapidly rotating solar-type stars, which differ markedly from the Sun.

astro-ph.SR↗

arXiv2025

TutorBench: A Benchmark To Assess Tutoring Capabilities Of Large Language Models

As students increasingly adopt large language models (LLMs) as learning aids, it is crucial to build models that are adept at handling the nuances of tutoring: they need to identify the core needs of students, be adaptive, provide personalized guidance, and be accurate. To this end, we introduce TutorBench, a dataset and evaluation benchmark designed to rigorously evaluate the core tutoring skills of LLMs. The dataset comprises 1,490 samples curated by human experts, focused on high-school and AP-level curricula. The samples are drawn from three common tutoring tasks: (i) generating adaptive explanations tailored to a student's confusion, (ii) providing actionable feedback on a student's work, and (iii) promoting active learning through effective hint generation. To account for the inherent complexity of tutoring, samples are accompanied by sample-specific rubrics which are used to judge model responses during evaluation. TutorBench uses a reliable and fine-grained automatic evaluation method that uses an LLM-judge and the sample-specific rubrics. We evaluate 16 frontier LLMs on TutorBench and present a detailed analysis of their performance and behavior. Our results show that none of the frontier LLMs achieve a score of greater than $56\%$, showing a large room for improvement. We find that LLMs fall short in exhibiting the full range of tutoring skills needed to guide, diagnose, and support students effectively, with all the frontier models achieving less than a $60\%$ pass rate on rubric criteria related to these skills. We also find that different model families exhibit varied strengths and limitations: the Claude models outperform others in supporting active learning, while they lag behind in the other two use cases. By releasing TutorBench, we provide a comprehensive and unsaturated benchmark to guide the development of the next-generation of AI tutors.

cs.LG↗

arXiv2025

MultiNRC: A Challenging and Native Multilingual Reasoning Evaluation Benchmark for LLMs

Although recent Large Language Models (LLMs) have shown rapid improvement on reasoning benchmarks in English, the evaluation of such LLMs' multilingual reasoning capability across diverse languages and cultural contexts remains limited. Existing multilingual reasoning benchmarks are typically constructed by translating existing English reasoning benchmarks, biasing these benchmarks towards reasoning problems with context in English language/cultures. In this work, we introduce the Multilingual Native Reasoning Challenge (MultiNRC), a benchmark designed to assess LLMs on more than 1,000 native, linguistic and culturally grounded reasoning questions written by native speakers in French, Spanish, and Chinese. MultiNRC covers four core reasoning categories: language-specific linguistic reasoning, wordplay & riddles, cultural/tradition reasoning, and math reasoning with cultural relevance. For cultural/tradition reasoning and math reasoning with cultural relevance, we also provide English equivalent translations of the multilingual questions by manual translation from native speakers fluent in English. This set of English equivalents can provide a direct comparison of LLM reasoning capacity in other languages vs. English on the same reasoning questions. We systematically evaluate current 14 leading LLMs covering most LLM families on MultiNRC and its English equivalent set. The results show that (1) current LLMs are still not good at native multilingual reasoning, with none scoring above 50% on MultiNRC; (2) LLMs exhibit distinct strengths and weaknesses in handling linguistic, cultural, and logical reasoning tasks; (3) Most models perform substantially better in math reasoning in English compared to in original languages (+10%), indicating persistent challenges with culturally grounded knowledge.

cs.CL↗

arXiv2025

Initiation Route of Coronal Mass Ejections: II. The Role of Filament Mass

The thorough understanding on the initiation of coronal mass ejections (CMEs), which is manifested as a slow rise of pre-eruptive structures before the impulsive ejection in kinematics, is the key for forecasting the solar eruptions. In our previous work, we showed that the slow rise of a hot flux rope with coronal mass density is caused by the moderate magnetic reconnection occurring in the hyperbolic flux tube (HFT) combined with the torus instability. However, it remains unclear how the initiation process varies when a filament is present in the pre-eruptive flux rope. In this work, we reveal the complete initiation route of a CME containing filament mass with a state-of-the-art full-magnetohydrodynamics simulation. The comprehensive analyses show that the filament mass has an important impact on the CME initiation through triggering and driving the slow rise of flux rope with its drainage, besides the contributions of HFT reconnection and torus instability. Finally, in combination with our previous work, we propose that the enhanced drainage of filament mass and various features related to the HFT reconnection, such as, the split of pre-eruptive structure and the pre-flare loops and X-ray emissions, can serve as the precursors of CME initiation in observations.

astro-ph.SR↗

arXiv2025

MultiFileTest: A Multi-File-Level LLM Unit Test Generation Benchmark and Impact of Error Fixing Mechanisms

Unit test generation has become a promising and important Large Language Model (LLM) use case. However, existing evaluation benchmarks for LLM unit test generation focus on function- or class-level code (single-file) rather than more practical and challenging multi-file-level codebases. To address such a limitation, we propose MultiFileTest, a multi-file-level benchmark for unit test generation covering Python, Java, and JavaScript. MultiFileTest features 20 moderate-sized and high-quality projects per language. We evaluate eleven frontier LLMs on MultiFileTest, and the results show that most frontier LLMs tested exhibit moderate performance on MultiFileTest, highlighting the difficulty of MultiFileTest. We also conduct a thorough error analysis, which shows that even advanced LLMs, such as Gemini-3.0-Pro, exhibit basic yet critical errors, including executability and cascade errors. Motivated by this observation, we further evaluate all frontier LLMs under manual error-fixing and self-error-fixing scenarios to assess their potential when equipped with error-fixing mechanisms. Our code and dataset is available at \href{https://github.com/YiboWANG214/ProjectTest}{MultiFileTest}.

cs.SE↗

arXiv2025

MultiChallenge: A Realistic Multi-Turn Conversation Evaluation Benchmark Challenging to Frontier LLMs

We present MultiChallenge, a pioneering benchmark evaluating large language models (LLMs) on conducting multi-turn conversations with human users, a crucial yet underexamined capability for their applications. MultiChallenge identifies four categories of challenges in multi-turn conversations that are not only common and realistic among current human-LLM interactions, but are also challenging to all current frontier LLMs. All 4 challenges require accurate instruction-following, context allocation, and in-context reasoning at the same time. We also develop LLM as judge with instance-level rubrics to facilitate an automatic evaluation method with fair agreement with experienced human raters. Despite achieving near-perfect scores on existing multi-turn evaluation benchmarks, all frontier models have less than 50% accuracy on MultiChallenge, with the top-performing Claude 3.5 Sonnet (June 2024) achieving just a 41.4% average accuracy.

cs.CL↗

arXiv2024

ReGenesis: LLMs can Grow into Reasoning Generalists via Self-Improvement

Post-training Large Language Models (LLMs) with explicit reasoning trajectories can enhance their reasoning abilities. However, acquiring such high-quality trajectory data typically demands meticulous supervision from humans or superior models, which can be either expensive or license-constrained. In this paper, we explore how far an LLM can improve its reasoning by self-synthesizing reasoning paths as training data without any additional supervision. Existing self-synthesizing methods, such as STaR, suffer from poor generalization to out-of-domain (OOD) reasoning tasks. We hypothesize it is due to that their self-synthesized reasoning paths are too task-specific, lacking general task-agnostic reasoning guidance. To address this, we propose Reasoning Generalist via Self-Improvement (ReGenesis), a method to self-synthesize reasoning paths as post-training data by progressing from abstract to concrete. More specifically, ReGenesis self-synthesizes reasoning paths by converting general reasoning guidelines into task-specific ones, generating reasoning structures, and subsequently transforming these structures into reasoning paths, without the need for human-designed task-specific examples used in existing methods. We show that ReGenesis achieves superior performance on all in-domain and OOD settings tested compared to existing methods. For six OOD tasks specifically, while previous methods exhibited an average performance decrease of approximately 4.6% after post training, ReGenesis delivers around 6.1% performance improvement. We also conduct in-depth analysis of our framework and show ReGenesis is effective across various LLMs and design choices.

cs.CL↗

arXiv2024

LLMs Assist NLP Researchers: Critique Paper (Meta-)Reviewing

This work is motivated by two key trends. On one hand, large language models (LLMs) have shown remarkable versatility in various generative tasks such as writing, drawing, and question answering, significantly reducing the time required for many routine tasks. On the other hand, researchers, whose work is not only time-consuming but also highly expertise-demanding, face increasing challenges as they have to spend more time reading, writing, and reviewing papers. This raises the question: how can LLMs potentially assist researchers in alleviating their heavy workload? This study focuses on the topic of LLMs assist NLP Researchers, particularly examining the effectiveness of LLM in assisting paper (meta-)reviewing and its recognizability. To address this, we constructed the ReviewCritique dataset, which includes two types of information: (i) NLP papers (initial submissions rather than camera-ready) with both human-written and LLM-generated reviews, and (ii) each review comes with "deficiency" labels and corresponding explanations for individual segments, annotated by experts. Using ReviewCritique, this study explores two threads of research questions: (i) "LLMs as Reviewers", how do reviews generated by LLMs compare with those written by humans in terms of quality and distinguishability? (ii) "LLMs as Metareviewers", how effectively can LLMs identify potential issues, such as Deficient or unprofessional review segments, within individual paper reviews? To our knowledge, this is the first work to provide such a comprehensive analysis.

cs.CL↗

arXiv2024

FOFO: A Benchmark to Evaluate LLMs' Format-Following Capability

This paper presents FoFo, a pioneering benchmark for evaluating large language models' (LLMs) ability to follow complex, domain-specific formats, a crucial yet underexamined capability for their application as AI agents. Despite LLMs' advancements, existing benchmarks fail to assess their format-following proficiency adequately. FoFo fills this gap with a diverse range of real-world formats and instructions, developed through an AI-Human collaborative method. Our evaluation across both open-source (e.g., Llama 2, WizardLM) and closed-source (e.g., GPT-4, PALM2, Gemini) LLMs highlights three key findings: open-source models significantly lag behind closed-source ones in format adherence; LLMs' format-following performance is independent of their content generation quality; and LLMs' format proficiency varies across different domains. These insights suggest the need for specialized tuning for format-following skills and highlight FoFo's role in guiding the selection of domain-specific AI agents. FoFo is released here at https://github.com/SalesforceAIResearch/FoFo.

cs.CL↗

arXiv2024

Unveiling the Initiation Route of Coronal Mass Ejections through their Slow Rise Phase

Understanding the early evolution of coronal mass ejections (CMEs), in particular their initiation, is the key to forecasting solar eruptions and induced disastrous space weather. Although many initiation mechanisms have been proposed, a full understanding of CME initiation, which is identified as a slow rise of CME progenitors in kinematics before the impulsive acceleration, remains elusive. Here, with a state-of-the-art thermal-magnetohydrodynamics simulation, we determine a complete CME initiation route in which multiple mainstream mechanisms occur in sequence yet are tightly coupled. The slow rise is first triggered and driven by the developing hyperbolic flux tube (HFT) reconnection. Subsequently, the slow rise continues as driven by the coupling of the HFT reconnection and the early development of torus instability. The end of the slow rise, i.e., the onset of the impulsive acceleration, is induced by the start of the fast magnetic reconnection coupled with the torus instability. These results unveil that the CME initiation is a complicated process involving multiple physical mechanisms, thus being hardly resolved by a single initiation mechanism.

astro-ph.SR↗

arXiv2023

GlueGen: Plug and Play Multi-modal Encoders for X-to-image Generation

Text-to-image (T2I) models based on diffusion processes have achieved remarkable success in controllable image generation using user-provided captions. However, the tight coupling between the current text encoder and image decoder in T2I models makes it challenging to replace or upgrade. Such changes often require massive fine-tuning or even training from scratch with the prohibitive expense. To address this problem, we propose GlueGen, which applies a newly proposed GlueNet model to align features from single-modal or multi-modal encoders with the latent space of an existing T2I model. The approach introduces a new training objective that leverages parallel corpora to align the representation spaces of different encoders. Empirical results show that GlueNet can be trained efficiently and enables various capabilities beyond previous state-of-the-art models: 1) multilingual language models such as XLM-Roberta can be aligned with existing T2I models, allowing for the generation of high-quality images from captions beyond English; 2) GlueNet can align multi-modal encoders such as AudioCLIP with the Stable Diffusion model, enabling sound-to-image generation; 3) it can also upgrade the current text encoder of the latent diffusion model for challenging case generation. By the alignment of various feature representations, the GlueNet allows for flexible and efficient integration of new functionality into existing T2I models and sheds light on X-to-image (X2I) generation.

cs.CV↗

arXiv2023

Identifying Footpoints of Pre-eruptive and Coronal Mass Ejection Flux Ropes with Sunspot Scars

The properties of pre-eruptive structures and coronal mass ejections (CMEs) are characterized by those of their footpoints, the latter of which thus attract great interest. However, how to identify the footpoints of pre-eruptive structures and how to identify the footpoints with ground-based instruments, still remain elusive. In this work, we study an arc-shaped structure intruding in the sunspot umbra. It is located close to the (pre-)eruptive flux rope footpoint and is thus expected to help identify the footpoint. We analyse this arc-shaped structure, which we name as "sunspot scar", in a CME event on 2012 July 12 and in two CME events in observationally-inspired MHD simulations performed by OHM and MPI-AMRVAC. The sunspot scar has a more inclined magnetic field with a weaker vertical component and a stronger horizontal component relative to that in the surrounding umbra and manifests as a light bridge in the white light passband. The hot field lines anchored in the sunspot scar are spatially at the transition between the flux rope and the background coronal loops, and temporally in the process of the slipping reconnection which builds up the flux rope. The sunspot scar and its related light bridge mark the edge of the CME flux rope footpoint, and especially, the edge of the pre-eruptive flux rope footpoint in the framework of "pre-eruptive structures being flux ropes". Therefore, they provide a new perspective for the identification of pre-eruptive and CME flux rope footpoints, and also new methods for studying the properties and evolution of pre-eruptive structures and CMEs with photospheric observations only.

astro-ph.SR↗

arXiv2023

Lemur: Harmonizing Natural Language and Code for Language Agents

We introduce Lemur and Lemur-Chat, openly accessible language models optimized for both natural language and coding capabilities to serve as the backbone of versatile language agents. The evolution from language chat models to functional language agents demands that models not only master human interaction, reasoning, and planning but also ensure grounding in the relevant environments. This calls for a harmonious blend of language and coding capabilities in the models. Lemur and Lemur-Chat are proposed to address this necessity, demonstrating balanced proficiencies in both domains, unlike existing open-source models that tend to specialize in either. Through meticulous pre-training using a code-intensive corpus and instruction fine-tuning on text and code data, our models achieve state-of-the-art averaged performance across diverse text and coding benchmarks among open-source models. Comprehensive experiments demonstrate Lemur's superiority over existing open-source models and its proficiency across various agent tasks involving human communication, tool usage, and interaction under fully- and partially- observable environments. The harmonization between natural and programming languages enables Lemur-Chat to significantly narrow the gap with proprietary models on agent abilities, providing key insights into developing advanced open-source agents adept at reasoning, planning, and operating seamlessly across environments. https://github.com/OpenLemur/Lemur

cs.CL↗

arXiv2023

XGen-7B Technical Report

Large Language Models (LLMs) have become ubiquitous across various domains, transforming the way we interact with information and conduct research. However, most high-performing LLMs remain confined behind proprietary walls, hindering scientific progress. Most open-source LLMs, on the other hand, are limited in their ability to support longer sequence lengths, which is a key requirement for many tasks that require inference over an input context. To address this, we have trained XGen, a series of 7B parameter models on up to 8K sequence length for up to 1.5T tokens. We have also finetuned the XGen models on public-domain instructional data, creating their instruction-tuned counterparts (XGen-Inst). We open-source our models for both research advancements and commercial applications. Our evaluation on standard benchmarks shows that XGen models achieve comparable or better results when compared with state-of-the-art open-source LLMs. Our targeted evaluation on long sequence modeling tasks shows the benefits of our 8K-sequence models over 2K-sequence open-source LLMs.

cs.CL↗