Text_Generation
-
TextPecker: Rewarding Structural Anomaly Quantification for Enhancing Visual Text Rendering
2026-02-24 13:40:23Hanshen Zhu, Yuliang Liu, Xuecheng Wu, An-Lan Wang, Hao Feng, Dingkang Yang, Chao Feng, Can Huang, Jingqun Tang, Xiang BaiAbstract
Visual Text Rendering (VTR) remains a critical challenge in text-to-image generation, where even advanced models frequently produce text with structural anomalies such as distortion, blurriness, and misalignment. However, we find that leading MLLMs and specialist OCR models largely fail to perceive these structural anomalies, creating a critical bottleneck for both VTR evaluation and RL-based optimization. As a result, even state-of-the-art generators (e.g., SeedDream4.0, Qwen-Image) still struggle to render structurally faithful text. To address this, we propose TextPecker, a plug-and-play structural anomaly perceptive RL strategy that mitigates noisy reward signals and works with any textto-image generator. To enable this capability, we construct a recognition dataset with character-level structural-anomaly annotations and develop a stroke-editing synthesis engine to expand structural-error coverage. Experiments show that TextPecker consistently improves diverse text-to-image models; even on the well-optimized Qwen-Image, it significantly yields average gains of 4% in structural fidelity and 8.7% in semantic alignment for Chinese text rendering, establishing a new state-of-the-art in high-fidelity VTR. Our work fills a gap in VTR optimization, providing a foundational step towards reliable and structural faithful visual text generation.
Abstract (translated)
视觉文本渲染（VTR）在文字到图像生成领域仍然是一个关键挑战，即使是最先进的模型也经常产生结构异常的文字，如扭曲、模糊和错位。然而，我们发现领先的多模态语言模型（MLLMs）和专业的OCR模型在识别这些结构异常方面表现不佳，这为VTR评估和基于强化学习的优化造成了重大瓶颈。因此，即使是最先进的生成器（例如SeedDream4.0 和 Qwen-Image），依然难以准确渲染出具有结构一致性的文字。 为了应对这一挑战，我们提出了TextPecker——一种即插即用的感知结构异常的强化学习策略，它可以减少噪声奖励信号，并与任何文本到图像生成模型兼容。为了实现这一点，我们构建了一个包含字符级别结构异常注释的数据集，并开发了一种笔画编辑合成引擎以扩展错误覆盖范围。 实验表明，TextPecker可以持续改进各种不同的文字到图像的模型；即使在已经高度优化的Qwen-Image上，它依然能够显著提升平均4%的文字结构保真度和8.7%的语义对齐度（针对中文文本渲染），确立了高保真VTR的新标杆。我们的工作填补了VTR优化中的空白，并为可靠且具有结构性一致性的视觉文字生成奠定了基础步骤。
URL
https://arxiv.org/abs/2602.20903
PDF
-
Agentic AI as a Cybersecurity Attack Surface: Threats, Exploits, and Defenses in Runtime Supply Chains
2026-02-23 06:57:57Xiaochong Jiang, Shiqi Yang, Wenting Yang, Yichen Liu, Cheng JiAbstract
Agentic systems built on large language models (LLMs) extend beyond text generation to autonomously retrieve information and invoke tools. This runtime execution model shifts the attack surface from build-time artifacts to inference-time dependencies, exposing agents to manipulation through untrusted data and probabilistic capability resolution. While prior work has focused on model-level vulnerabilities, security risks emerging from cyclic and interdependent runtime behavior remain fragmented. We systematize these risks within a unified runtime framework, categorizing threats into data supply chain attacks (transient context injection and persistent memory poisoning) and tool supply chain attacks (discovery, implementation, and invocation). We further identify the Viral Agent Loop, in which agents act as vectors for self-propagating generative worms without exploiting code-level flaws. Finally, we advocate a Zero-Trust Runtime Architecture that treats context as untrusted control flow and constrains tool execution through cryptographic provenance rather than semantic inference.
Abstract (translated)
基于大型语言模型（LLM）的代理系统不仅限于文本生成，还能自主检索信息并调用工具。这种运行时执行模式将攻击面从构建时的工件转移到了推理时的依赖关系上，使代理更容易受到通过不可信数据和概率能力解析进行操纵的影响。尽管之前的工作主要集中在模型层面的安全漏洞上，但源自循环及相互依存的运行行为所出现的安全风险依然分散且未系统化。我们在此框架内对这些风险进行了系统化的梳理，并将威胁分为两类：数据供应链攻击（瞬态上下文注入和持久内存中毒）以及工具供应链攻击（发现、实现与调用）。此外，我们还识别出“病毒代理循环”，在这种情况下，即使不利用代码层面的漏洞，代理也能作为自我传播生成型蠕虫的载体。最后，我们倡导采用零信任运行时架构，在这种架构中，将上下文视为不可信的控制流，并通过加密来源而非语义推理来限制工具执行。
URL
https://arxiv.org/abs/2602.19555
PDF
-
MICON-Bench: Benchmarking and Enhancing Multi-Image Context Image Generation in Unified Multimodal Models
2026-02-23 04:32:52Mingrui Wu, Hang Liu, Jiayi Ji, Xiaoshuai Sun, Rongrong JiAbstract
Recent advancements in Unified Multimodal Models (UMMs) have enabled remarkable image understanding and generation capabilities. However, while models like Gemini-2.5-Flash-Image show emerging abilities to reason over multiple related images, existing benchmarks rarely address the challenges of multi-image context generation, focusing mainly on text-to-image or single-image editing tasks. In this work, we introduce \textbf{MICON-Bench}, a comprehensive benchmark covering six tasks that evaluate cross-image composition, contextual reasoning, and identity preservation. We further propose an MLLM-driven Evaluation-by-Checkpoint framework for automatic verification of semantic and visual consistency, where multimodal large language model (MLLM) serves as a verifier. Additionally, we present \textbf{Dynamic Attention Rebalancing (DAR)}, a training-free, plug-and-play mechanism that dynamically adjusts attention during inference to enhance coherence and reduce hallucinations. Extensive experiments on various state-of-the-art open-source models demonstrate both the rigor of MICON-Bench in exposing multi-image reasoning challenges and the efficacy of DAR in improving generation quality and cross-image coherence. Github: this https URL.
Abstract (translated)
最近在统一多模态模型（UMMs）方面的发展使得图像理解和生成能力显著提升。然而，尽管像Gemini-2.5-Flash-Image这样的模型展示了处理多个相关图片的能力，现有的基准测试却很少关注跨多张图片的上下文生成挑战，主要集中在文本到图像或单张图片编辑任务上。在本研究中，我们引入了\textbf{MICON-Bench}，这是一个全面的基准测试，涵盖了评估跨图象组合、上下文推理和身份保留这六个任务。此外，我们还提出了一种由多模态大型语言模型（MLLM）驱动的通过检查点进行自动验证框架——称为\textbf{MLLM-Driven Evaluation-by-Checkpoint}，用于检测语义和视觉一致性。进一步地，我们提出了一个名为\textbf{动态注意重新平衡（DAR）}的方法，这是一个无需训练且即插即用的机制，在推理过程中动态调整注意力权重以增强连贯性和减少幻觉现象。 在多种最先进的开源模型上的广泛实验表明，MICON-Bench能够严格揭示多图象推理中的挑战，并展示了DAR方法在提升生成质量和跨图像一致性方面的有效性。Github地址：this https URL.
URL
https://arxiv.org/abs/2602.19497
PDF
-
DoAtlas-1: A Causal Compilation Paradigm for Clinical AI
2026-02-22 12:40:52Yulong Li, Jianxu Chen, Xiwei Liu, Chuanyue Suo, Rong Xia, Zhixiang Lu, Yichen Li, Xinlin Zhuang, Niranjana Arun Menon, Yutong Xie, Eran Segal, Imran RazzakAbstract
Medical foundation models generate narrative explanations but cannot quantify intervention effects, detect evidence conflicts, or validate literature claims, limiting clinical auditability. We propose causal compilation, a paradigm that transforms medical evidence from narrative text into executable code. The paradigm standardizes heterogeneous research evidence into structured estimand objects, each explicitly specifying intervention contrast, effect scale, time horizon, and target population, supporting six executable causal queries: do-calculus, counterfactual reasoning, temporal trajectories, heterogeneous effects, mechanistic decomposition, and joint interventions. We instantiate this paradigm in DoAtlas-1, compiling 1,445 effect kernels from 754 studies through effect standardization, conflict-aware graph construction, and real-world validation (Human Phenotype Project, 10,000 participants). The system achieves 98.5% canonicalization accuracy and 80.5% query executability. This paradigm shifts medical AI from text generation to executable, auditable, and verifiable causal reasoning.
Abstract (translated)
医学基础模型可以生成叙述性解释，但无法量化干预效果、检测证据冲突或验证文献声明，这限制了临床审计的可行性。我们提出因果编译（causal compilation）的概念，这是一种将医疗证据从叙述文本转化为可执行代码的范式。该方法通过标准化异质研究证据为结构化的估计对象，每个对象明确指明干预对比、影响规模、时间范围和目标人群，支持六种可执行的因果查询：do-演算（do-calculus）、反事实推理、时间轨迹分析、异质性效应分析、机制分解以及联合干预。我们通过DoAtlas-1实现了这一范式，在这个系统中编译了来自754项研究的1,445个效果内核，这些编译过程包括效果标准化、冲突感知图构建和现实世界验证（人类表型项目，涉及10,000名参与者）。该系统的规范性准确率为98.5%，查询执行率为80.5%。这一范式将医学人工智能从文本生成转向了可执行的、可审计的和可验证的因果推理。
URL
https://arxiv.org/abs/2602.19158
PDF
-
IDLM: Inverse-distilled Diffusion Language Models
2026-02-22 06:47:04David Li, Nikita Gushchin, Dmitry Abulkhanov, Eric Moulines, Ivan Oseledets, Maxim Panov, Alexander KorotinAbstract
Diffusion Language Models (DLMs) have recently achieved strong results in text generation. However, their multi-step sampling leads to slow inference, limiting practical use. To address this, we extend Inverse Distillation, a technique originally developed to accelerate continuous diffusion models, to the discrete setting. Nonetheless, this extension introduces both theoretical and practical challenges. From a theoretical perspective, the inverse distillation objective lacks uniqueness guarantees, which may lead to suboptimal solutions. From a practical standpoint, backpropagation in the discrete space is non-trivial and often unstable. To overcome these challenges, we first provide a theoretical result demonstrating that our inverse formulation admits a unique solution, thereby ensuring valid optimization. We then introduce gradient-stable relaxations to support effective training. As a result, experiments on multiple DLMs show that our method, Inverse-distilled Diffusion Language Models (IDLM), reduces the number of inference steps by 4x-64x, while preserving the teacher model's entropy and generative perplexity.
Abstract (translated)
扩散语言模型（DLM）最近在文本生成方面取得了显著成果。然而，其多步骤采样导致推理速度慢，限制了其实用性。为了解决这个问题，我们将逆向蒸馏技术扩展到离散场景中使用。这种技术最初是为了加速连续扩散模型而开发的。但是，这一扩展带来了理论和实践上的挑战。从理论上讲，逆向蒸馏目标缺乏唯一性保证，可能导致次优解出现。在实践中，离散空间中的反向传播复杂且不稳定。 为了克服这些挑战，我们首先提供了一个理论结果，证明我们的逆向公式具有唯一的解决方案，从而确保了有效的优化过程。接着，我们引入了梯度稳定的松弛方法来支持有效训练。实验结果显示，在多个DLM上应用这种方法——逆蒸馏扩散语言模型（IDLM）之后，推理步骤减少了4到64倍，并且保持了教师模型的熵和生成困惑度。 这种改进使得在不牺牲性能的情况下加速文本生成过程成为可能，从而提高了模型的实际使用效率。
URL
https://arxiv.org/abs/2602.19066
PDF
-
Learning to Detect Language Model Training Data via Active Reconstruction
2026-02-22 03:20:06Junjie Oscar Yin, John X. Morris, Vitaly Shmatikov, Sewon Min, Hannaneh HajishirziAbstract
Detecting LLM training data is generally framed as a membership inference attack (MIA) problem. However, conventional MIAs operate passively on fixed model weights, using log-likelihoods or text generations. In this work, we introduce \textbf{Active Data Reconstruction Attack} (ADRA), a family of MIA that actively induces a model to reconstruct a given text through training. We hypothesize that training data are \textit{more reconstructible} than non-members, and the difference in their reconstructibility can be exploited for membership inference. Motivated by findings that reinforcement learning (RL) sharpens behaviors already encoded in weights, we leverage on-policy RL to actively elicit data reconstruction by finetuning a policy initialized from the target model. To effectively use RL for MIA, we design reconstruction metrics and contrastive rewards. The resulting algorithms, \textsc{ADRA} and its adaptive variant \textsc{ADRA+}, improve both reconstruction and detection given a pool of candidate data. Experiments show that our methods consistently outperform existing MIAs in detecting pre-training, post-training, and distillation data, with an average improvement of 10.7\% over the previous runner-up. In particular, \MethodPlus~improves over Min-K\%++ by 18.8\% on BookMIA for pre-training detection and by 7.6\% on AIME for post-training detection.
Abstract (translated)
检测大型语言模型（LLM）的训练数据通常被视为成员推断攻击（MIA）问题。然而，传统的MIA是被动操作固定模型权重的方法，使用对数似然或文本生成。在这项工作中，我们引入了**主动数据重建攻击**（ADRA），这是一个通过训练诱导模型重构给定文本的一类MIA方法。我们的假设是，训练数据比非成员数据更具可重构性，并且它们之间可重构性的差异可以用于成员推断。鉴于强化学习（RL）能够增强权重中已编码的行为这一发现，我们利用在目标模型上初始化策略的在策略RL来主动引发数据重建。为了有效使用RL进行MIA，我们设计了重构指标和对比奖励。由此产生的算法，ADRA及其自适应变体ADRA+，在给定候选数据池的情况下，在重构和检测方面都得到了改进。实验表明，我们的方法在检测预训练、后训练以及蒸馏数据时，相对于现有的MIA方法平均提高了10.7%，超越了之前的亚军。特别地，\MethodPlus 在BookMIA上的预训练检测中比Min-K\%++提升了18.8%，而在AIME上的后训练检测中提升了7.6%。
URL
https://arxiv.org/abs/2602.19020
PDF
-
Whisper: Courtside Edition Enhancing ASR Performance Through LLM-Driven Context Generation
2026-02-21 22:15:59Yonathan Ron, Shiri Gilboa, Tammuz DubnovAbstract
Domain-specific speech remains a persistent challenge for automatic speech recognition (ASR), even for state-of-the-art systems like OpenAI's Whisper. We introduce Whisper: Courtside Edition, a novel multi-agent large language model (LLM) pipeline that enhances Whisper transcriptions without retraining. The pipeline intercepts Whisper's initial transcript, applies specialized LLM agents for domain context identification, named entity recognition, and jargon detection, and generates compact prompts that guide Whisper's decoder. Evaluated on 421 NBA basketball commentary segments (a domain characterized by dense proper nouns and technical terminology) our best pipeline achieves a statistically significant 17.0% relative reduction in word error rate (WER; from 0.217 to 0.180, p<0.001). Improvements are observed in 40.1% of segments with degradation in only 7.1%, substantially outperforming direct transcript post-editing. These results demonstrate that prompt-based augmentation can deliver scalable domain adaptation for ASR, offering a practical alternative to costly model fine-tuning.
Abstract (translated)
领域特定的语音识别仍然是自动语音识别（ASR）系统面临的一个持久挑战，即使对于像OpenAI的Whisper这样的最先进的系统也是如此。我们引入了Whisper：Courtside Edition，这是一个新颖的多代理大型语言模型（LLM）管道，可以在不重新训练的情况下增强Whisper的转录效果。该管道拦截Whisper生成的初始转录，并应用专门的LLM代理进行领域上下文识别、命名实体识别和行话检测，然后生成简短提示来指导Whisper解码器的工作。 在评估中，我们使用了421段NBA篮球评论（这一领域以密集的人名和其他专有名词以及技术术语为特征）测试我们的最佳管道。结果表明，在统计学上有显著意义的情况下，相对错误率降低了17.0%（从0.217降至0.180，p<0.001）。在40.1%的片段中观察到了改进，只有7.1%的片段出现退化现象。这远远优于直接对转录进行后编辑的效果。 这些结果表明基于提示的增强技术可以为ASR提供可扩展领域的适应性，并且作为一种替代方案，相对于成本高昂的模型微调来说更加实际可行。
URL
https://arxiv.org/abs/2602.18966
PDF
-
Evo: Autoregressive-Diffusion Large Language Models with Evolving Balance
2026-02-20 15:26:52Junde Wu, Minhao Hu, Jiayuan Zhu, Yuyuan Liu, Tianyi Zhang, Kang Li, Jingkun Chen, Jiazhen Pan, Min Xu, Yueming JinAbstract
We introduce \textbf{Evo}, a duality latent trajectory model that bridges autoregressive (AR) and diffusion-based language generation within a continuous evolutionary generative framework. Rather than treating AR decoding and diffusion generation as separate paradigms, Evo reconceptualizes text generation as a latent flow: each token is associated with a vector-valued embedding that evolves over a progression variable $t_i \in [0, 1]$, indicating its semantic maturity. Low $t_i$ values correspond to confident AR-like refinement, while high values invoke diffusion-style planning, allowing the model to adaptively balance AR and diffusion based on uncertainty. Theoretically, we show that both AR and diffusion models emerge as discretizations of a shared probability flow, and we derive Evo's training objective from a unified variational ELBO. The model is implemented as a time-conditioned Transformer governed by a shared vector field, trained end-to-end to jointly infer latent codes and their progression times. During decoding, Evo performs efficient, semantics-aware refinement, achieving high-quality outputs without sacrificing speed. Empirically, Evo 8B achieves state-of-the-art or highly competitive results on 15 diverse benchmarks, including reasoning (GSM8K, ARC-C), code generation (HumanEval, MBPP), and general language understanding, while maintaining fast inference speed. Our results demonstrate that Evo delivers a new paradigm for LLM design with strong generation quality, robust symbolic reasoning, and decoding efficiency.
Abstract (translated)
我们介绍了一种新的模型——\textbf{Evo}，它是一种连续进化生成框架中的二元潜在轨迹模型，用于连接自回归（AR）和基于扩散的语言生成。Evo 不是将 AR 解码和扩散生成视为独立的范式，而是重新构思文本生成为一个潜在流：每个词与一个矢量值嵌入相关联，该嵌入随时间变量 $t_i \in [0, 1]$ 进化，表示其语义成熟度。低 $t_i$ 值对应于自信的 AR 类似精炼过程，而高值则激发扩散风格规划，使模型能够根据不确定性灵活地在自回归和基于扩散的方法之间进行切换。 从理论上讲，我们展示了自回归和扩散模型都是共享概率流的一种离散化，并且我们从统一的变分ELBO中导出了Evo的训练目标。该模型通过一个受时间条件限制并由共同矢量场控制的Transformer实现，在端到端训练过程中同时推理潜在代码及其进度时间。 在解码阶段，Evo 实现了高效的、语义感知的精炼过程，可以在不牺牲速度的情况下生成高质量输出。实验表明，Evo 8B 在包括逻辑推理（GSM8K, ARC-C）、代码生成（HumanEval, MBPP）和通用语言理解在内的15个多样化基准测试中取得了当前最佳或极具竞争力的结果，并且保持了快速的推断速度。 我们的结果证明，Evo 提供了一种全新的大型语言模型设计范式，具有强大的生成质量、稳健的符号推理能力和高效的解码效率。
URL
https://arxiv.org/abs/2603.06617
PDF
-
MusicSem: A Semantically Rich Language--Audio Dataset of Natural Music Descriptions
2026-02-19 19:01:42Rebecca Salganik, Teng Tu, Fei-Yueh Chen, Xiaohao Liu, Keifeng Lu, Ethan Luvisia, Zhiyao Duan, Guillaume Salha-Galvan, Anson Kahng, Yunshan Ma, Jian KangAbstract
Music representation learning is central to music information retrieval and generation. While recent advances in multimodal learning have improved alignment between text and audio for tasks such as cross-modal music retrieval, text-to-music generation, and music-to-text generation, existing models often struggle to capture users' expressed intent in natural language descriptions of music. This observation suggests that the datasets used to train and evaluate these models do not fully reflect the broader and more natural forms of human discourse through which music is described. In this paper, we introduce MusicSem, a dataset of 32,493 language-audio pairs derived from organic music-related discussions on the social media platform Reddit. Compared to existing datasets, MusicSem captures a broader spectrum of musical semantics, reflecting how listeners naturally describe music in nuanced and human-centered ways. To structure these expressions, we propose a taxonomy of five semantic categories: descriptive, atmospheric, situational, metadata-related, and contextual. In addition to the construction, analysis, and release of MusicSem, we use the dataset to evaluate a wide range of multimodal models for retrieval and generation, highlighting the importance of modeling fine-grained semantics. Overall, MusicSem serves as a novel semantics-aware resource to support future research on human-aligned multimodal music representation learning.
Abstract (translated)
音乐表示学习是音乐信息检索和生成的核心。虽然近期在多模态学习方面的进展提升了文本与音频之间的对齐能力，特别是在跨模式音乐检索、从文本到音乐的生成以及从音乐到文本的生成等任务中，现有的模型往往难以捕捉用户在其自然语言描述中的意图表达。这一观察表明，用于训练和评估这些模型的数据集未能充分反映人类通过广泛且自然的语言形式来描述音乐的方式。 在本文中，我们介绍了MusicSem数据集，该数据集包含从社交媒体平台Reddit上收集的32,493对语言-音频配对，这些内容源自有机的音乐相关讨论。与现有的数据集相比，MusicSem捕捉到了更广泛的音乐语义范围，反映了听众如何以细腻且以人为中心的方式自然地描述音乐。为了结构化这些表达，我们提出了一个包含五种语义类别的分类法：描述性、氛围性、情境性、元数据相关的和上下文性的。 除了构建、分析并发布MusicSem之外，我们还利用该数据集评估了广泛的多模态模型在检索和生成任务中的表现，并强调了建模细粒度语义的重要性。总体而言，MusicSem作为一个新颖的语义感知资源，旨在支持未来关于与人类行为一致的多模态音乐表示学习的研究工作。
URL
https://arxiv.org/abs/2602.17769
PDF
-
From Labor to Collaboration: A Methodological Experiment Using AI Agents to Augment Research Perspectives in Taiwan's Humanities and Social Sciences
Abstract
Generative AI is reshaping knowledge work, yet existing research focuses predominantly on software engineering and the natural sciences, with limited methodological exploration for the humanities and social sciences. Positioned as a "methodological experiment," this study proposes an AI Agent-based collaborative research workflow (Agentic Workflow) for humanities and social science research. Taiwan's this http URL usage data (N = 7,729 conversations, November 2025) from the Anthropic Economic Index (AEI) serves as the empirical vehicle for validating the feasibility of this methodology. This study operates on two levels: the primary level is the design and validation of a methodological framework - a seven-stage modular workflow grounded in three principles: task modularization, human-AI division of labor, and verifiability, with each stage delineating clear roles for human researchers (research judgment and ethical decisions) and AI Agents (information retrieval and text generation); the secondary level is the empirical analysis of AEI Taiwan data - serving as an operational demonstration of the workflow's application to secondary data research, showcasing both the process and output quality (see Appendix A). This study contributes by proposing a replicable AI collaboration framework for humanities and social science researchers, and identifying three operational modes of human-AI collaboration - direct execution, iterative refinement, and human-led - through reflexive documentation of the operational process. This taxonomy reveals the irreplaceability of human judgment in research question formulation, theoretical interpretation, contextualized reasoning, and ethical reflection. Limitations including single-platform data, cross-sectional design, and AI reliability risks are acknowledged.
Abstract (translated)
URL
https://arxiv.org/abs/2602.17221
PDF
-
AREG: Adversarial Resource Extraction Game for Evaluating Persuasion and Resistance in Large Language Models
2026-02-18 17:28:28Adib Sakhawat, Fardeen SadabAbstract
Evaluating the social intelligence of Large Language Models (LLMs) increasingly requires moving beyond static text generation toward dynamic, adversarial interaction. We introduce the Adversarial Resource Extraction Game (AREG), a benchmark that operationalizes persuasion and resistance as a multi-turn, zero-sum negotiation over financial resources. Using a round-robin tournament across frontier models, AREG enables joint evaluation of offensive (persuasion) and defensive (resistance) capabilities within a single interactional framework. Our analysis provides evidence that these capabilities are weakly correlated ($\rho = 0.33$) and empirically dissociated: strong persuasive performance does not reliably predict strong resistance, and vice versa. Across all evaluated models, resistance scores exceed persuasion scores, indicating a systematic defensive advantage in adversarial dialogue settings. Further linguistic analysis suggests that interaction structure plays a central role in these outcomes. Incremental commitment-seeking strategies are associated with higher extraction success, while verification-seeking responses are more prevalent in successful defenses than explicit refusal. Together, these findings indicate that social influence in LLMs is not a monolithic capability and that evaluation frameworks focusing on persuasion alone may overlook asymmetric behavioral vulnerabilities.
Abstract (translated)
URL
https://arxiv.org/abs/2602.16639
PDF
-
Aladdin-FTI @ AMIYA Three Wishes for Arabic NLP: Fidelity, Diglossia, and Multidialectal Generation
2026-02-18 09:15:20Jonathan Mutal, Perla Al Almaoui, Simon Hengchen, Pierrette BouillonAbstract
Arabic dialects have long been under-represented in Natural Language Processing (NLP) research due to their non-standardization and high variability, which pose challenges for computational modeling. Recent advances in the field, such as Large Language Models (LLMs), offer promising avenues to address this gap by enabling Arabic to be modeled as a pluricentric language rather than a monolithic system. This paper presents Aladdin-FTI, our submission to the AMIYA shared task. The proposed system is designed to both generate and translate dialectal Arabic (DA). Specifically, the model supports text generation in Moroccan, Egyptian, Palestinian, Syrian, and Saudi dialects, as well as bidirectional translation between these dialects, Modern Standard Arabic (MSA), and English. The code and trained model are publicly available.
Abstract (translated)
URL
https://arxiv.org/abs/2602.16290
PDF
-
SpreadsheetArena: Decomposing Preference in LLM Generation of Spreadsheet Workbooks
2026-02-16 14:24:36Srivatsa Kundurthy, Clara Na, Michael Handley, Zach Kirshner, Chen Bo Calvin Zhang, Manasi Sharma, Emma Strubell, John LingAbstract
Large language models (LLMs) are increasingly tasked with producing and manipulating structured artifacts. We consider the task of end-to-end spreadsheet generation, where language models are prompted to produce spreadsheet artifacts to satisfy users' explicit and implicit constraints, specified in natural language. We introduce SpreadsheetArena, a platform for evaluating models' performance on the task via blind pairwise evaluations of LLM-generated spreadsheet workbooks. As with other complex, open-ended tasks, relevant evaluation criteria can vary substantially across use cases and prompts, often in ways that are difficult to formalize. Compared to general chat or text generation settings, spreadsheet generation presents unique challenges and opportunities: the task output structure is well-defined and multi-dimensional, and there are often complex considerations around interactivity and layout. Among other findings, we observe that stylistic, structural, and functional features of preferred spreadsheets vary substantially across use cases, and expert evaluations of spreadsheets for finance prompts suggests that even highly ranked arena models do not reliably produce spreadsheets aligned with domain-specific best practices. Our hope is that our work prompts further study of end-to-end spreadsheet generation as a challenging and interesting category of complex, open-ended tasks for LLMs. Our live arena is hosted at this https URL.
Abstract (translated)
大型语言模型（LLMs）越来越多地被用于生成和处理结构化数据。我们研究了一个端到端的电子表格生成任务，在这个任务中，语言模型根据自然语言提示来生成满足用户明确或隐含需求的电子表格。为此，我们引入了SpreadsheetArena平台，这是一个通过盲对评估（blind pairwise evaluations）来评价LLM生成电子表格工作簿表现的平台。 与其他复杂的开放式任务一样，相关的评判标准可能会因使用案例和提示的不同而有很大差异，且这些标准往往难以形式化。与一般的聊天或文本生成设置相比，电子表格生成具有独特的挑战性和机遇：其输出结构是明确且多维度的，并且常常存在复杂的人机交互以及布局方面的考量。 我们的研究发现，不同应用场景中用户偏好电子表格的风格、结构和功能特征差异显著；而对于财务相关提示的专家评估则表明，即使是排名靠前的模型生成的电子表格也未必符合特定领域的最佳实践。我们希望这项工作能够激发更多关于端到端电子表格生成作为复杂开放式任务领域中的挑战性话题的研究。 我们的在线竞赛平台可在此 [链接]（请将"this https URL"替换为实际的有效URL）访问。
URL
https://arxiv.org/abs/2603.10002
PDF
-
CLASE: A Hybrid Method for Chinese Legalese Stylistic Evaluation
2026-02-13 05:51:56Yiran Rex Ma, Yuxiao Ye, Huiyuan XieAbstract
Legal text generated by large language models (LLMs) can usually achieve reasonable factual accuracy, but it frequently fails to adhere to the specialised stylistic norms and linguistic conventions of legal writing. In order to improve stylistic quality, a crucial first step is to establish a reliable evaluation method. However, having legal experts manually develop such a metric is impractical, as the implicit stylistic requirements in legal writing practice are difficult to formalise into explicit rubrics. Meanwhile, existing automatic evaluation methods also fall short: reference-based metrics conflate semantic accuracy with stylistic fidelity, and LLM-as-a-judge evaluations suffer from opacity and inconsistency. To address these challenges, we introduce CLASE (Chinese LegAlese Stylistic Evaluation), a hybrid evaluation method that focuses on the stylistic performance of legal text. The method incorporates a hybrid scoring mechanism that combines 1) linguistic feature-based scores and 2) experience-guided LLM-as-a-judge scores. Both the feature coefficients and the LLM scoring experiences are learned from contrastive pairs of authentic legal documents and their LLM-restored counterparts. This hybrid design captures both surface-level features and implicit stylistic norms in a transparent, reference-free manner. Experiments on 200 Chinese legal documents show that CLASE achieves substantially higher alignment with human judgments than traditional metrics and pure LLM-as-a-judge methods. Beyond improved alignment, CLASE provides interpretable score breakdowns and suggestions for improvements, offering a scalable and practical solution for professional stylistic evaluation in legal text generation (Code and data for CLASE is available at: this https URL).
Abstract (translated)
URL
https://arxiv.org/abs/2602.12639
PDF
-
Decoder-only Conformer with Modality-aware Sparse Mixtures of Experts for ASR
2026-02-13 02:53:54Jaeyoung Lee, Masato MimuraAbstract
We present a decoder-only Conformer for automatic speech recognition (ASR) that processes speech and text in a single stack without external speech encoders or pretrained large language models (LLM). The model uses a modality-aware sparse mixture of experts (MoE): disjoint expert pools for speech and text with hard routing and top-1 selection, embedded in hybrid-causality Conformer blocks (bidirectional for speech, causal for text). Training combines CTC on speech positions with label-smoothed cross-entropy for text generation. Our 113M-parameter model consistently improves WER over a 139M AED baseline on Librispeech (2.8% vs. 3.2% test-clean; 5.6% vs. 6.0% test-other). On Common Voice 16.1 with a single multilingual model across five languages, our approach reduces average WER from 12.2% to 10.6%. To our knowledge, this is the first randomly initialized decoder-only ASR that surpasses strong AED baselines via modality-aware routing and sparse MoE, achieving better accuracy with fewer active parameters and without alignment/adaptation modules.
Abstract (translated)
URL
https://arxiv.org/abs/2602.12546
PDF
-
Autorubric: A Unified Framework for Rubric-Based LLM Evaluation
2026-02-13 02:26:30Delip Rao, Chris Callison-BurchAbstract
Rubric-based evaluation with large language models (LLMs) has become standard practice for assessing text generation at scale, yet the underlying techniques are scattered across papers with inconsistent terminology and partial solutions. We present a unified framework: each identified technique is paired with its realization in Autorubric, an open-source Python framework proposed in this paper. Autorubric supports binary, ordinal, and nominal criteria with configurable weights; single-judge and multi-judge ensemble evaluation with majority, weighted, unanimous, and any-vote aggregation; few-shot calibration with verdict-balanced sampling; and mitigations for position bias (option shuffling), verbosity bias (length penalties), and criterion conflation (per-criterion atomic evaluation with natural language explanations). The framework provides reliability metrics drawn from psychometrics (Cohen's $\kappa$, weighted $\kappa$, correlation coefficients, and distribution-level tests) alongside production infrastructure including response caching, checkpointing with resumable runs, multi-provider rate limiting, and cost tracking. We evaluate Autorubric on three benchmarks spanning educational assessment, deep research evaluation, and chatbot quality assessment, demonstrating that it produces results consistent with published benchmarks while exercising the framework's key capabilities: per-criterion binary evaluation with few-shot calibration (RiceChem), multi-judge ensemble evaluation across judge models (ResearcherBench), and mixed criterion types combining binary, ordinal, and nominal scales (CHARM-100). We also contribute CHARM-100, a 100-sample chatbot evaluation dataset with per-sample ground truth labels across all three criterion types, designed to stress-test rubric evaluation frameworks on heterogeneous criteria.
Abstract (translated)
基于评分标准的大规模语言模型（LLMs）评估已成为大规模文本生成评估的标准实践，但其底层技术分散在不同的论文中，并且术语不一致、解决方案也不完整。本文提出了一种统一的框架：每一种被识别的技术都与其在Autorubric中的实现相配对，Autorubric是本文提出的开源Python框架。该框架支持二元、序数和名义标准，并可以配置权重；单评价员和多评价员集成评估包括多数投票、加权投票、一致同意和任意投票聚合方法；少量样本校准采用判决平衡抽样；并且有缓解位置偏差（选项洗牌）、冗长偏见（长度惩罚）以及评判混淆的措施（每个评判点原子级评估并附带自然语言解释）。该框架提供了来自心理测量学的可靠性指标（Cohen's κ、加权κ、相关系数和分布水平测试），同时包括了生产基础设施，如响应缓存、可恢复运行的检查点、多提供商速率限制以及成本跟踪。我们通过三个涵盖教育评估、深度研究评估及聊天机器人质量评估的基准来评测Autorubric，证明它能产生与已发布基准一致的结果，同时也展示了其关键功能：基于少量样本校准的单评判标准二元评价（RiceChem），跨越不同评判模型的多评价员集成评价（ResearcherBench），以及结合了二元、序数和名义尺度的混合评判类型评估（CHARM-100）。我们还贡献了一个名为CHARM-100的数据集，它包含了100个样本的聊天机器人评估数据，并为所有三种评判标准提供了每个样本的真实标签，旨在考验评分标准评价框架在异构准则上的性能。
URL
https://arxiv.org/abs/2603.00077
PDF
-
T3D: Few-Step Diffusion Language Models via Trajectory Self-Distillation with Direct Discriminative Optimization
2026-02-12 18:52:35Tunyu Zhang, Xinxi Zhang, Ligong Han, Haizhou Shi, Xiaoxiao He, Zhuowei Li, Hao Wang, Kai Xu, Akash Srivastava, Hao Wang, Vladimir Pavlovic, Dimitris N. MetaxasAbstract
Diffusion large language models (DLLMs) have the potential to enable fast text generation by decoding multiple tokens in parallel. However, in practice, their inference efficiency is constrained by the need for many refinement steps, while aggressively reducing the number of steps leads to a substantial degradation in generation quality. To alleviate this, we propose a trajectory self-distillation framework that improves few-step decoding by distilling the model's own generative trajectories. We incorporate Direct Discriminative Optimization (DDO), a reverse-KL objective that promotes mode-seeking distillation and encourages the student to concentrate on high-probability teacher modes. Across benchmarks, our approach consistently outperforms strong few-step baselines and standard training under tight step budgets. Although full-step decoding remains superior, we substantially narrow the gap, establishing a strong foundation towards practical few-step DLLMs. The source code is available at this https URL.
Abstract (translated)
扩散型大型语言模型（DLLM）具有通过并行解码多个标记来实现快速文本生成的潜力。然而，在实践中，其推理效率受到需要多次细化步骤的限制，而过于激进地减少这些步骤会导致生成质量大幅下降。为了缓解这一问题，我们提出了一种轨迹自我蒸馏框架，该框架通过提炼模型自身的生成轨迹来改进少步解码。我们引入了直接判别性优化（DDO），这是一种反向KL目标函数，它促进了模式寻求的蒸馏，并鼓励学生集中在高概率教师模式上。在各个基准测试中，我们的方法始终优于强大的少步基线和标准训练下的紧缩步骤预算。尽管全步解码仍然更优，但我们显著缩小了差距，为实用的少步DLLM建立了坚实的基础。源代码可在该链接获取。
URL
https://arxiv.org/abs/2602.12262
PDF
-
When Should LLMs Be Less Specific? Selective Abstraction for Reliable Long-Form Text Generation
2026-02-12 13:06:14Shani Goren, Ido Galil, Ran El-YanivAbstract
LLMs are widely used, yet they remain prone to factual errors that erode user trust and limit adoption in high-risk settings. One approach to mitigate this risk is to equip models with uncertainty estimation mechanisms that abstain when confidence is low. However, this binary "all-or-nothing" approach is excessively restrictive in long-form settings, often discarding valuable information. We introduce Selective Abstraction (SA), a framework that enables LLMs to trade specificity for reliability by selectively reducing the detail of uncertain content. We first formalize SA through the lenses of selective risk and coverage. We then propose Atom-wise Selective Abstraction, a claim-level instantiation that decomposes responses into atomic claims (short, self-contained statements each expressing a single fact) and replaces uncertain atoms with higher confidence, less specific abstractions. To evaluate this framework, we develop a novel end-to-end pipeline for open-ended generation that instantiates risk as factual correctness and measures coverage using an information-theoretic measure of retained information. Across six open-source models on the FactScore and LongFact-Objects benchmarks, atom-wise SA consistently outperforms existing baselines, improving the area under the risk-coverage curve (AURC) by up to 27.73% over claim removal, demonstrating that reducing specificity can boost accuracy and reliability while preserving most of their original meaning.
Abstract (translated)
大型语言模型（LLMs）被广泛应用，但仍容易出现事实性错误，这会削弱用户信任并限制其在高风险环境中的应用。一种缓解这一问题的方法是为模型配备不确定性估计机制，在信心不足时拒绝生成内容。然而，在长文本场景中这种“要么全有要么全无”的方法过于严格，往往导致有价值的资讯被舍弃。为此，我们引入了选择性抽象（Selective Abstraction, SA）框架，该框架允许LLMs通过选择性减少不确定内容的细节来在具体性和可靠性之间进行权衡。 首先，我们从选择风险和覆盖率的角度对SA进行了形式化定义。然后，提出了原子级选择性抽象（Atom-wise Selective Abstraction），这是一种基于声明级别的实现方式，它将响应分解成原子声明（即每一个表达单一事实的简短、独立的信息单元），并用更高置信度但更不具体的抽象信息替换不确定的原子。 为了评估这一框架的有效性，我们开发了一种新颖的端到端生成管道，在这种管道中，风险被定义为事实准确性，而覆盖率则通过一种基于信息论的方法来衡量保留下来的信息量。在六款开源模型上，针对FactScore和LongFact-Objects基准测试数据集进行了实验，原子级SA始终优于现有基线方法，并将风险-覆盖曲线下面积（AURC）提高了最多27.73%，优于声明删除法的表现。这表明减少具体性可以在保持大部分原始含义的同时提高准确性和可靠性。
URL
https://arxiv.org/abs/2602.11908
PDF
-
Prototype Transformer: Towards Language Model Architectures Interpretable by Design
2026-02-12 11:43:39Yordan Yordanov, Matteo Forasassi, Bayar Menzat, Ruizhi Wang, Chang Qi, Markus Kaltenberger, Amine M'Charrak, Tommaso Salvatori, Thomas LukasiewiczAbstract
While state-of-the-art language models (LMs) surpass the vast majority of humans in certain domains, their reasoning remains largely opaque, undermining trust in their output. Furthermore, while autoregressive LMs can output explicit reasoning, their true reasoning process is opaque, which introduces risks like deception and hallucination. In this work, we introduce the Prototype Transformer (ProtoT) -- an autoregressive LM architecture based on prototypes (parameter vectors), posed as an alternative to the standard self-attention-based transformers. ProtoT works by means of two-way communication between the input sequence and the prototypes, and we show that this leads to the prototypes automatically capturing nameable concepts (e.g. "woman") during training. They provide the potential to interpret the model's reasoning and allow for targeted edits of its behavior. Furthermore, by design, the prototypes create communication channels that aggregate contextual information at different time scales, aiding interpretability. In terms of computation scalability, ProtoT scales linearly with sequence length vs the quadratic scalability of SOTA self-attention transformers. Compared to baselines, ProtoT scales well with model and data size, and performs well on text generation and downstream tasks (GLUE). ProtoT exhibits robustness to input perturbations on par or better than some baselines, but differs from them by providing interpretable pathways showing how robustness and sensitivity arises. Reaching close to the performance of state-of-the-art architectures, ProtoT paves the way to creating well-performing autoregressive LMs interpretable by design.
Abstract (translated)
尽管最先进的语言模型（LMs）在某些领域超越了绝大多数人类，但它们的推理过程仍然很大程度上不透明，这削弱了人们对这些模型输出的信任。此外，虽然自回归LM能够生成明确的推理，但其真正的推理过程依然是不透明的，从而引入了诸如欺骗和幻觉等风险。在这项工作中，我们介绍了原型变压器（ProtoT）——一种基于原型（参数向量）的自回归LM架构，并将其作为标准基于自我注意机制的transformer的一个替代方案。 ProtoT通过输入序列与原型之间的双向通信工作，在训练过程中，它能够自动捕获可命名的概念（例如“女性”），从而提供了解释模型推理的可能性，并允许对模型行为进行有针对性的修改。此外，从设计上讲，这些原型创建了聚合不同时间尺度上下文信息的沟通渠道，有助于提高模型的可解释性。 在计算扩展性方面，ProtoT随序列长度线性扩展，而最先进的自注意transformer则呈二次方扩展。与基准相比，ProtoT在模型和数据规模上扩展良好，并且在文本生成和下游任务（GLUE）中表现出色。ProtoT对输入扰动的鲁棒性表现可媲美或优于某些基准模型，但其独特之处在于提供了解释路径，展示了如何产生这种鲁棒性和敏感性的。 达到接近最先进架构性能的同时，ProtoT为创建具有内在可解释性的高性能自回归LM开辟了道路。
URL
https://arxiv.org/abs/2602.11852
PDF
-
RELATE: A Reinforcement Learning-Enhanced LLM Framework for Advertising Text Generation
2026-02-12 10:00:55Jinfang Wang, Jiajie Liu, Jianwei Wu, Ziqin Luo, Zhen Chen, Chunlei Li, Biao Han, Tao Deng, Yi Li, Shuanglong Li, Lin LiuAbstract
In online advertising, advertising text plays a critical role in attracting user engagement and driving advertiser value. Existing industrial systems typically follow a two-stage paradigm, where candidate texts are first generated and subsequently aligned with online performance metrics such as click-through rate(CTR). This separation often leads to misaligned optimization objectives and low funnel efficiency, limiting global optimality. To address these limitations, we propose RELATE, a reinforcement learning-based end-to-end framework that unifies generation and objective alignment within a single model. Instead of decoupling text generation from downstream metric alignment, RELATE integrates performance and compliance objectives directly into the generation process via policy learning. To better capture ultimate advertiser value beyond click-level signals, We incorporate conversion-oriented metrics into the objective and jointly model them with compliance constraints as multi-dimensional rewards, enabling the model to generate high-quality ad texts that improve conversion performance under policy constraints. Extensive experiments on large-scale industrial datasets demonstrate that RELATE consistently outperforms baselines. Furthermore, online deployment on a production advertising platform yields statistically significant improvements in click-through conversion rate(CTCVR) under strict policy constraints, validating the robustness and real-world effectiveness of the proposed framework.
Abstract (translated)
在在线广告领域，广告文案对于吸引用户互动和增加广告主价值起着关键作用。现有的工业系统通常遵循两阶段模式：首先是生成候选文本，然后将这些文本与点击率（CTR）等线上性能指标进行对齐。这种分离往往导致优化目标不一致及低效率的漏斗效应，从而限制了全局最优性。为了解决这些问题，我们提出了一种基于强化学习的端到端框架——RELATE，它通过单一模型统一生成和目标对齐过程。与将文本生成从下游指标对齐中分离不同，RELATE 通过策略学习直接将性能和合规性目标融入生成过程中。 为了更好地捕捉超出点击级信号的最终广告价值，我们引入了面向转化率的目标，并将其与合规约束共同建模为多维奖励，使模型能够在政策限制下生成高质量且能提升转化表现的广告文案。大规模工业数据集上的广泛实验表明，RELATE 一直优于基线方法。此外，在一个生产广告平台上进行的在线部署在严格的政策条件下显著提高了点击后转化率（CTCVR），验证了所提框架的稳健性和实际效果。
URL
https://arxiv.org/abs/2602.11780
PDF
