Text_Generation
-
eTracer: Towards Traceable Text Generation via Claim-Level Grounding
2026-01-07 07:44:30Bohao Chu, Qianli Wang, Hendrik Damm, Hui Wang, Ula Muhabbek, Elisabeth Livingstone, Christoph M. Friedrich, Norbert FuhrAbstract
How can system-generated responses be efficiently verified, especially in the high-stakes biomedical domain? To address this challenge, we introduce eTracer, a plug-and-play framework that enables traceable text generation by grounding claims against contextual evidence. Through post-hoc grounding, each response claim is aligned with contextual evidence that either supports or contradicts it. Building on claim-level grounding results, eTracer not only enables users to precisely trace responses back to their contextual source but also quantifies response faithfulness, thereby enabling the verifiability and trustworthiness of generated responses. Experiments show that our claim-level grounding approach alleviates the limitations of conventional grounding methods in aligning generated statements with contextual sentence-level evidence, resulting in substantial improvements in overall grounding quality and user verification efficiency. The code and data are available at this https URL.
Abstract (translated)
如何高效验证系统生成的响应，尤其是在高风险的生命科学领域？为应对这一挑战，我们引入了eTracer，这是一个即插即用框架，通过将声明与上下文证据进行对齐来实现可追溯的文本生成。通过事后对齐方式，每个响应声明都会与其支持或反驳它的上下文证据相匹配。基于声明级别的对齐结果，eTracer不仅使用户能够准确地追踪响应到其上下文来源，而且还量化了响应的真实性，从而实现了生成响应的验证性和可信度。实验表明，我们提出的声明级别对齐方法缓解了传统对齐方法在将生成语句与上下文句子级证据相匹配方面的局限性，这极大地提高了整体对齐质量和用户验证效率。代码和数据可在以下链接获取：[此URL]（请将此处的"this https URL"替换为实际提供的有效链接）。
URL
https://arxiv.org/abs/2601.03669
PDF
-
FlowPlan-G2P: A Structured Generation Framework for Transforming Scientific Papers into Patent Descriptions
Abstract
Over 3.5 million patents are filed annually, with drafting patent descriptions requiring deep technical and legal expertise. Transforming scientific papers into patent descriptions is particularly challenging due to their differing rhetorical styles and stringent legal requirements. Unlike black-box text-to-text approaches that struggle to model structural reasoning and legal constraints, we propose FlowPlan-G2P, a novel framework that mirrors the cognitive workflow of expert drafters by reformulating this task into three stages: (1) Concept Graph Induction, extracting technical entities and relationships into a directed graph via expert-like reasoning; (2) Paragraph and Section Planning, reorganizing the graph into coherent clusters aligned with canonical patent sections; and (3) Graph-Conditioned Generation, producing legally compliant paragraphs using section-specific subgraphs and tailored prompts. Experiments demonstrate that FlowPlan-G2P significantly improves logical coherence and legal compliance over end-to-end LLM baselines. Our framework establishes a new paradigm for paper-to-patent generation and advances structured text generation for specialized domains.
Abstract (translated)
每年有超过350万项专利申请，撰写专利描述需要深厚的技术和法律专业知识。将科技论文转化为专利说明书尤其具有挑战性，这主要是因为它们在修辞风格上存在差异，并且要满足严格的法律规定。不同于那些难以建模结构化推理及法律约束的黑箱文本到文本的方法，我们提出了FlowPlan-G2P这一全新框架，它通过模仿专家撰稿人的认知工作流程，将任务分解为三个阶段来处理： 1. **概念图诱导**：利用类似专家的推理方式，从科技论文中提取技术实体和关系，并将其组织成一个有向图。 2. **段落与章节规划**：重新排列这些节点和边以形成一系列相互关联且连贯的小簇，并将它们对齐到标准专利章节之下。 3. **条件图生成**：使用特定于每个章节的子图以及定制提示，生成符合法律规定的段落。 实验表明，FlowPlan-G2P在逻辑一致性和法律合规性方面相比端到端LLM基准有了显著提高。我们所提出的框架为论文转专利文本生成设立了新的范式，并推动了专门领域内的结构化文本生成技术的发展。
URL
https://arxiv.org/abs/2601.02589
PDF
-
ModeX: Evaluator-Free Best-of-N Selection for Open-Ended Generation
2026-01-05 20:16:32Hyeong Kyu Choi, Sharon LiAbstract
Selecting a single high-quality output from multiple stochastic generations remains a fundamental challenge for large language models (LLMs), particularly in open-ended tasks where no canonical answer exists. While Best-of-N and self-consistency methods show that aggregating multiple generations can improve performance, existing approaches typically rely on external evaluators, reward models, or exact string-match voting, limiting their applicability and efficiency. We propose Mode Extraction (ModeX), an evaluator-free Best-of-N selection framework that generalizes majority voting to open-ended text generation by identifying the modal output representing the dominant semantic consensus among generated texts. ModeX constructs a similarity graph over candidate generations and recursively applies spectral clustering to select a representative centroid, without requiring additional inference or auxiliary models. We further instantiate this selection principle as ModeX--Lite, an improved version of ModeX with early pruning for efficiency. Across open-ended tasks--including text summarization, code generation, and mathematical reasoning--our approaches consistently outperform standard single- and multi-path baselines, providing a computationally efficient solution for robust open-ended text generation. Code is released in this https URL.
Abstract (translated)
从多个随机生成的结果中选择一个高质量的输出，仍然是大型语言模型（LLM）面临的基本挑战，尤其是在没有标准答案存在的开放性任务中。虽然Best-of-N和自我一致性方法表明聚合多次生成可以提高性能，但现有的方法通常依赖于外部评估者、奖励模型或精确字符串匹配投票，这限制了它们的应用范围和效率。我们提出了模式提取（ModeX），这是一种无需评估器的Best-of-N选择框架，通过识别生成文本中代表主导语义共识的模态输出，将多数表决推广到了开放性文本生成任务中。ModeX构建了一个候选生成之间的相似性图，并递归地应用谱聚类来选择一个代表性中心点，而不需要额外的推理或辅助模型。我们进一步实现为ModeX-Lite，这是模式提取的一个改进版本，通过早期剪枝提高了效率。在包括文本摘要、代码生成和数学推理在内的开放性任务中，我们的方法始终优于标准单路径或多路径基线，提供了一种计算上高效且稳健的解决方案用于开放性文本生成。相关代码可以在提供的URL地址获取。
URL
https://arxiv.org/abs/2601.02535
PDF
-
Deferred Commitment Decoding for Diffusion Language Models with Confidence-Aware Sliding Windows
2026-01-05 12:57:33Yingte Shu, Yuchuan Tian, Chao Xu, Yunhe Wang, Hanting ChenAbstract
Diffusion language models (DLMs) have recently emerged as a strong alternative to autoregressive models by enabling parallel text generation. To improve inference efficiency and KV-cache compatibility, prior work commonly adopts block-based diffusion, decoding tokens block by block. However, this paradigm suffers from a structural limitation that we term Boundary-Induced Context Truncation (BICT): undecoded tokens near block boundaries are forced to commit without access to nearby future context, even when such context could substantially reduce uncertainty. This limitation degrades decoding confidence and generation quality, especially for tasks requiring precise reasoning, such as mathematical problem solving and code generation. We propose Deferred Commitment Decoding (DCD), a novel, training-free decoding strategy that mitigates this issue. DCD maintains a confidence-aware sliding window over masked tokens, resolving low-uncertainty tokens early while deferring high-uncertainty tokens until sufficient contextual evidence becomes available. This design enables effective bidirectional information flow within the decoding window without sacrificing efficiency. Extensive experiments across multiple diffusion language models, benchmarks, and caching configurations show that DCD improves generation accuracy by 1.39% with comparable time on average compared to fixed block-based diffusion methods, with the most significant improvement reaching 9.0%. These results demonstrate that deferring token commitment based on uncertainty is a simple yet effective principle for improving both the quality and efficiency of diffusion language model decoding.
Abstract (translated)
最近，扩散语言模型（DLMs）作为自回归模型的有力替代者而兴起，它们通过允许并行文本生成来实现这一点。为了提高推理效率和KV缓存兼容性，先前的工作通常采用基于块的扩散方法，即分块解码令牌。然而，这种范式存在一个我们称之为边界诱导上下文截断（BICT）的结构性限制：接近块边界的未解码令牌被迫在没有访问到附近未来上下文的情况下做出承诺，即使这样的上下文可以显著减少不确定性。这一限制降低了解码信心和生成质量，尤其是在需要精确推理的任务中，例如数学问题求解和代码生成。 我们提出了一种新的、无需训练的解码策略——延迟承诺解码（DCD），以缓解这个问题。DCD在屏蔽令牌上维护一个基于信心的滑动窗口，早期解决低不确定性令牌，而高不确定性的令牌则被推迟直到有足够的上下文证据再做出决定。这种设计使得解码窗口内的双向信息流得以有效进行，而不牺牲效率。 跨多个扩散语言模型、基准和缓存配置的广泛实验表明，在与固定块式扩散方法的时间成本相当的情况下，DCD平均将生成准确性提高了1.39%，最高可达9%的显著改进。这些结果证明了根据不确定性延迟令牌承诺是一种简单而有效的原则，可以提高扩散语言模型解码的质量和效率。
URL
https://arxiv.org/abs/2601.02076
PDF
-
MORE: Multi-Objective Adversarial Attacks on Speech Recognition
2026-01-05 07:27:57Xiaoxue Gao, Zexin Li, Yiming Chen, Nancy F. ChenAbstract
The emergence of large-scale automatic speech recognition (ASR) models such as Whisper has greatly expanded their adoption across diverse real-world applications. Ensuring robustness against even minor input perturbations is therefore critical for maintaining reliable performance in real-time environments. While prior work has mainly examined accuracy degradation under adversarial attacks, robustness with respect to efficiency remains largely unexplored. This narrow focus provides only a partial understanding of ASR model vulnerabilities. To address this gap, we conduct a comprehensive study of ASR robustness under multiple attack scenarios. We introduce MORE, a multi-objective repetitive doubling encouragement attack, which jointly degrades recognition accuracy and inference efficiency through a hierarchical staged repulsion-anchoring mechanism. Specifically, we reformulate multi-objective adversarial optimization into a hierarchical framework that sequentially achieves the dual objectives. To further amplify effectiveness, we propose a novel repetitive encouragement doubling objective (REDO) that induces duplicative text generation by maintaining accuracy degradation and periodically doubling the predicted sequence length. Overall, MORE compels ASR models to produce incorrect transcriptions at a substantially higher computational cost, triggered by a single adversarial input. Experiments show that MORE consistently yields significantly longer transcriptions while maintaining high word error rates compared to existing baselines, underscoring its effectiveness in multi-objective adversarial attack.
Abstract (translated)
大规模自动语音识别（ASR）模型如Whisper的出现，极大地扩展了其在各种现实世界应用中的采用。因此，在面对轻微输入变化时确保模型鲁棒性对于保持实时环境下的可靠性能至关重要。尽管先前的工作主要研究了对抗攻击下准确率下降的问题，但关于效率方面的鲁棒性研究仍然相对较少。这种狭隘的关注点仅提供了对ASR模型脆弱性的部分理解。为了解决这一差距，我们进行了一项多场景攻击情况下的ASR模型鲁棒性全面研究。 为此，我们引入了MORE（多目标重复加倍鼓励攻击），该攻击通过分层阶段排斥-锚定机制同时降低识别准确率和推理效率。具体而言，我们将一个多目标对抗优化问题重新表述为一个逐步实现双重目标的层次框架。为了进一步放大效果，我们提出了一种新颖的重复鼓励加倍目标(REDO)，通过保持准确性下降的同时周期性地将预测序列长度加倍来诱导生成冗余文本。 总的来说，MORE促使ASR模型以更高的计算成本在单次对抗输入触发下产生错误转录。实验表明，与现有基线相比，MORE能够持续输出显著更长的转录，同时保持高词错率，这证明了其在多目标对抗攻击中的有效性。
URL
https://arxiv.org/abs/2601.01852
PDF
-
QSLM: A Performance- and Memory-aware Quantization Framework with Tiered Search Strategy for Spike-driven Language Models
2026-01-02 13:05:33Rachmad Vidya Wicaksana Putra, Pasindu Wickramasinghe, Muhammad ShafiqueAbstract
Large Language Models (LLMs) have been emerging as prominent AI models for solving many natural language tasks due to their high performance (e.g., accuracy) and capabilities in generating high-quality responses to the given inputs. However, their large computational cost, huge memory footprints, and high processing power/energy make it challenging for their embedded deployments. Amid several tinyLLMs, recent works have proposed spike-driven language models (SLMs) for significantly reducing the processing power/energy of LLMs. However, their memory footprints still remain too large for low-cost and resource-constrained embedded devices. Manual quantization approach may effectively compress SLM memory footprints, but it requires a huge design time and compute power to find the quantization setting for each network, hence making this approach not-scalable for handling different networks, performance requirements, and memory budgets. To bridge this gap, we propose QSLM, a novel framework that performs automated quantization for compressing pre-trained SLMs, while meeting the performance and memory constraints. To achieve this, QSLM first identifies the hierarchy of the given network architecture and the sensitivity of network layers under quantization, then employs a tiered quantization strategy (e.g., global-, block-, and module-level quantization) while leveraging a multi-objective performance-and-memory trade-off function to select the final quantization setting. Experimental results indicate that our QSLM reduces memory footprint by up to 86.5%, reduces power consumption by up to 20%, maintains high performance across different tasks (i.e., by up to 84.4% accuracy of sentiment classification on the SST-2 dataset and perplexity score of 23.2 for text generation on the WikiText-2 dataset) close to the original non-quantized model while meeting the performance and memory constraints.
Abstract (translated)
大型语言模型（LLMs）由于其高性能和生成高质量响应的能力，已成为解决许多自然语言任务的突出AI模型。然而，它们巨大的计算成本、庞大的内存占用以及对处理能力和能源的巨大需求，使其在嵌入式设备上的部署变得具有挑战性。鉴于此，最近的研究提出了一种称为脉冲驱动语言模型（SLMs）的方法，以显著降低LLM的处理功率和能耗。尽管如此，这些模型仍然因内存占用过大而难以应用于低成本、资源受限的嵌入式设备中。 手动量化方法可以通过压缩SLM的内存足迹来有效解决问题，但这种方法需要大量的设计时间和计算能力来为每个网络找到合适的量化设置，因此对于不同网络、性能需求和内存预算而言并不具备可扩展性。为了弥合这一差距，我们提出了QSLM——一种新的框架，它能够自动执行预训练SLMs的量化以压缩其大小，同时满足性能和内存约束。 为了实现这个目标，QSLM首先识别给定网络架构的层次结构以及在量化过程中各层的敏感性，然后采用分层量化策略（例如全局、块级和模块级量化），并利用一个兼顾性能与内存双重优化的目标函数来选择最终的量化设置。实验结果显示，我们的QSLM可以在满足性能和内存约束的同时，将模型的内存占用减少多达86.5%，能耗降低最多达20%。同时，在不同的任务上保持高精度（例如SST-2数据集情感分类准确率为84.4%，WikiText-2数据集文本生成的困惑度得分为23.2），接近未量化模型的表现水平。
URL
https://arxiv.org/abs/2601.00679
PDF
-
PackKV: Reducing KV Cache Memory Footprint through LLM-Aware Lossy Compression
2025-12-30 20:05:32Bo Jiang, Taolue Yang, Youyuan Liu, Xubin He, Sheng Di, Sian JinAbstract
Transformer-based large language models (LLMs) have demonstrated remarkable potential across a wide range of practical applications. However, long-context inference remains a significant challenge due to the substantial memory requirements of the key-value (KV) cache, which can scale to several gigabytes as sequence length and batch size increase. In this paper, we present \textbf{PackKV}, a generic and efficient KV cache management framework optimized for long-context generation. %, which synergistically supports both latency-critical and throughput-critical inference scenarios. PackKV introduces novel lossy compression techniques specifically tailored to the characteristics of KV cache data, featuring a careful co-design of compression algorithms and system architecture. Our approach is compatible with the dynamically growing nature of the KV cache while preserving high computational efficiency. Experimental results show that, under the same and minimum accuracy drop as state-of-the-art quantization methods, PackKV achieves, on average, \textbf{153.2}\% higher memory reduction rate for the K cache and \textbf{179.6}\% for the V cache. Furthermore, PackKV delivers extremely high execution throughput, effectively eliminating decompression overhead and accelerating the matrix-vector multiplication operation. Specifically, PackKV achieves an average throughput improvement of \textbf{75.7}\% for K and \textbf{171.7}\% for V across A100 and RTX Pro 6000 GPUs, compared to cuBLAS matrix-vector multiplication kernels, while demanding less GPU memory bandwidth. Code available on this https URL
Abstract (translated)
翻译： 基于Transformer的大型语言模型（LLMs）已经在广泛的实用应用中展示了非凡的潜力。然而，长上下文推理仍然是一个重大挑战，因为关键值（KV）缓存的记忆需求随着序列长度和批量大小的增长而显著增加，可能达到数GB之多。在本文中，我们提出了\textbf{PackKV}，这是一个专为长上下文生成优化的通用且高效的KV缓存管理框架。%该方法同时支持延迟敏感型和吞吐量敏感型推理场景。 PackKV引入了专门针对KV缓存数据特性的新颖有损压缩技术，并巧妙地将压缩算法与系统架构相结合设计。我们的方法可以与动态增长性质的KV缓存兼容，同时保持高效的计算效率。实验结果表明，在相同的或最小化的精度下降情况下，与最先进的量化方法相比，PackKV在K缓存中平均实现了\textbf{153.2%}更高的内存减少率，在V缓存中则为\textbf{179.6%}。 此外，PackKV能够实现极高的执行吞吐量，有效消除解压缩开销并加速矩阵向量乘法操作。具体来说，与cuBLAS矩阵向量乘核相比，PackKV在A100和RTX Pro 6000 GPU上分别实现了K平均\textbf{75.7%}和V平均\textbf{171.7%}的吞吐量改进，并且所需的GPU内存带宽更少。 代码可在以下网址获取：[请在此处插入实际链接]
URL
https://arxiv.org/abs/2512.24449
PDF
-
Path Integral Solution for Dissipative Generative Dynamics
Abstract
Can purely mechanical systems generate intelligent language? We prove that dissipative quantum dynamics with analytically tractable non-local context aggregation produce coherent text generation, while conservation laws cause fundamental failure. Employing Koopman operators with closed-form path integral propagators, we show irreversible computation fundamentally requires both controlled information dissipation and causal context aggregation. Spectral analysis reveals emergent eigenvalue structure, separating into decay modes (forgetting), growth modes (amplification), and neutral modes (preservation) -- the essential ingredients for directed information flow. Hamiltonian constraints force the elimination of these dissipative modes and degrading performance despite unchanged model capacity. This establishes language generation as dissipative quantum field theory, proving mechanical systems acquire intelligence through the combination of dissipation and non-locality, not through conservation.
Abstract (translated)
完全机械系统能否生成智能语言？我们证明，具有解析可处理的非局部上下文聚合的动力耗散量子动力学会产生连贯的文本生成，而守恒定律则会导致根本性的失败。通过使用具有封闭形式路径积分传播器的Koopman算子，我们展示了不可逆计算从根本上需要受控信息耗散和因果上下文聚合。谱分析揭示了出现的特征值结构，分为衰减模式（遗忘）、增长模式（放大）和中性模式（保存），这些是定向信息流的关键成分。哈密顿量约束强制消除了这些耗散模式，并且即使模型容量不变，性能也会下降。这将语言生成确立为耗散量子场论，证明了机械系统通过耗散和非局部性的结合获得智能，而不是通过守恒。
URL
https://arxiv.org/abs/2601.00860
PDF
-
GRADE: Replacing Policy Gradients with Backpropagation for LLM Alignment
2025-12-30 03:45:32Lukas Abrie NelAbstract
Reinforcement learning from human feedback (RLHF) has become the dominant paradigm for aligning large language models with human preferences. However, policy gradient methods such as PPO suffer from high variance gradient estimates, requiring careful hyperparameter tuning and extensive computational resources. We introduce GRADE (Gumbel-softmax Relaxation for Alignment via Differentiable Estimation), a method that replaces high-variance policy gradient estimation with direct backpropagation through a differentiable relaxation of the discrete token sampling process. Using the Gumbel-Softmax reparameterization with straight-through estimation (GRADE-STE), we enable end-to-end gradient flow from reward signals through generated tokens to model parameters. On sentiment-controlled text generation using the IMDB dataset, GRADE-STE achieves a test reward of 0.763 +- 0.344 compared to PPO's 0.510 +- 0.313 and REINFORCE's 0.617 +- 0.378, representing a 50% relative improvement over PPO. Critically, GRADE-STE exhibits gradient variance over 14 times lower than REINFORCE and maintains stable training dynamics throughout optimization. Our rigorous evaluation with proper train/validation/test splits demonstrates that these improvements generalize to held-out data, with GRADE-STE showing the best generalization characteristics among all methods tested. GRADE offers a simpler, more stable, and more effective alternative to reinforcement learning for LLM alignment.
Abstract (translated)
从人类反馈中的强化学习（RLHF）已成为将大型语言模型与人类偏好对齐的主导方法。然而，策略梯度方法如PPO（近端策略优化）由于高方差的梯度估计而需要仔细调整超参数和大量的计算资源。 我们提出了一种名为GRADE（通过可微估计进行对齐的Gumbel-softmax松弛）的方法，该方法通过直接反向传播穿过离散标记采样过程的一个可微分松弛来替代高方差策略梯度估计。使用带有直通估计（GRADE-STE）的Gumbel-Softmax参数化，我们使从奖励信号到生成令牌再到模型参数的端到端梯度流动成为可能。 在使用IMDB数据集进行情感控制文本生成时，GRADE-STE实现了0.763 ± 0.344的测试奖励值，相比之下PPO为0.510 ± 0.313，REINFORCE为0.617 ± 0.378，这代表了与PPO相比约50%的相对改进。更重要的是，GRADE-STE表现出比REINFORCE低14倍以上的梯度方差，并在整个优化过程中保持稳定的训练动态。 我们通过适当的训练/验证/测试拆分进行了严格的评估，这些结果表明这些改进可以推广到未见过的数据上，在所有测试的方法中，GRADE-STE显示出了最佳的泛化特性。GRADE为大型语言模型对齐提供了一种更为简单、稳定和有效的替代强化学习方法。
URL
https://arxiv.org/abs/2601.11574
PDF
-
A Stepwise-Enhanced Reasoning Framework for Large Language Models Based on External Subgraph Generation
2025-12-29 10:35:53Xin Zhang, Yang Cao, Baoxing Wu, Xinyi Chen, Kai Song, Siying LiAbstract
Large Language Models (LLMs) have achieved strong performance across a wide range of natural language processing tasks in recent years, including machine translation, text generation, and question answering. As their applications extend to increasingly complex scenarios, however, LLMs continue to face challenges in tasks that require deep reasoning and logical inference. In particular, models trained on large scale textual corpora may incorporate noisy or irrelevant information during generation, which can lead to incorrect predictions or outputs that are inconsistent with factual knowledge. To address this limitation, we propose a stepwise reasoning enhancement framework for LLMs based on external subgraph generation, termed SGR. The proposed framework dynamically constructs query relevant subgraphs from external knowledge bases and leverages their semantic structure to guide the reasoning process. By performing reasoning in a step by step manner over structured subgraphs, SGR reduces the influence of noisy information and improves reasoning accuracy. Specifically, the framework first generates an external subgraph tailored to the input query, then guides the model to conduct multi step reasoning grounded in the subgraph, and finally integrates multiple reasoning paths to produce the final answer. Experimental results on multiple benchmark datasets demonstrate that SGR consistently outperforms strong baselines, indicating its effectiveness in enhancing the reasoning capabilities of LLMs.
Abstract (translated)
近年来，大型语言模型（LLM）在包括机器翻译、文本生成和问答等众多自然语言处理任务中取得了优异的成绩。然而，在其应用扩展到更复杂场景时，这些模型仍然面临着需要深度推理和逻辑推断的任务的挑战。具体而言，基于大规模文本语料库训练的模型可能会在生成过程中吸收噪声或不相关的信息，导致预测错误或与事实知识不符的结果。为了应对这一限制，我们提出了一种基于外部子图生成的逐步推理增强框架SGR。该框架能够动态地从外部知识库中构建查询相关的子图，并利用其语义结构来引导推理过程。通过在结构化的子图上进行分步推理，SGR减少了噪声信息的影响并提高了推理准确性。 具体来说，该框架首先为输入的查询生成一个定制化的外部子图，然后指导模型基于此子图执行多步骤推理，最后整合多个推理路径以产生最终答案。实验结果表明，在多个基准数据集上的测试中，SGR始终优于强大的基线方法，这证明了其在增强LLM推理能力方面的有效性。
URL
https://arxiv.org/abs/2512.23356
PDF
-
Chinese Morph Resolution in E-commerce Live Streaming Scenarios
2025-12-29 08:04:48Jiahao Zhu, Jipeng Qiang, Ran Bai, Chenyu Liu, Xiaoye OuyangAbstract
E-commerce live streaming in China, particularly on platforms like Douyin, has become a major sales channel, but hosts often use morphs to evade scrutiny and engage in false advertising. This study introduces the Live Auditory Morph Resolution (LiveAMR) task to detect such violations. Unlike previous morph research focused on text-based evasion in social media and underground industries, LiveAMR targets pronunciation-based evasion in health and medical live streams. We constructed the first LiveAMR dataset with 86,790 samples and developed a method to transform the task into a text-to-text generation problem. By leveraging large language models (LLMs) to generate additional training data, we improved performance and demonstrated that morph resolution significantly enhances live streaming regulation.
Abstract (translated)
在中国，特别是在像抖音这样的平台上，电商直播已成为一个重要的销售渠道。然而，主播们经常使用变音（或称“换声音”）来逃避监管并进行虚假宣传。这项研究引入了“Live Auditory Morph Resolution (LiveAMR)”任务，旨在检测此类违规行为。与之前主要关注社交媒体和地下行业中的文本规避的形态变化研究不同，LiveAMR专注于健康和医疗直播中基于发音的规避。 为了开展这一研究，我们构建了第一个LiveAMR数据集，其中包括86,790个样本，并开发了一种方法将任务转化为一个从语音到文本生成的问题。通过利用大型语言模型（LLMs）来生成额外的训练数据，我们的性能得到了显著提升，并证明形态分析在很大程度上增强了对直播监管的有效性。 简而言之，这项研究通过对健康和医疗领域的电商直播进行发音识别，开发了一种新的检测虚假宣传的方法，并通过大规模的语言模型辅助技术提升了该方法的效果。
URL
https://arxiv.org/abs/2512.23280
PDF
-
BeHGAN: Bengali Handwritten Word Generation from Plain Text Using Generative Adversarial Networks
2025-12-25 14:38:12Md. Rakibul Islam, Md. Kamrozzaman Bhuiyan, Safwan Muntasir, Arifur Rahman Jawad, Most. Sharmin Sultana SamuAbstract
Handwritten Text Recognition (HTR) is a well-established research area. In contrast, Handwritten Text Generation (HTG) is an emerging field with significant potential. This task is challenging due to the variation in individual handwriting styles. A large and diverse dataset is required to generate realistic handwritten text. However, such datasets are difficult to collect and are not readily available. Bengali is the fifth most spoken language in the world. While several studies exist for languages such as English and Arabic, Bengali handwritten text generation has received little attention. To address this gap, we propose a method for generating Bengali handwritten words. We developed and used a self-collected dataset of Bengali handwriting samples. The dataset includes contributions from approximately five hundred individuals across different ages and genders. All images were pre-processed to ensure consistency and quality. Our approach demonstrates the ability to produce diverse handwritten outputs from input plain text. We believe this work contributes to the advancement of Bengali handwriting generation and can support further research in this area.
Abstract (translated)
手写文本识别（HTR）是一个研究领域，已经相当成熟。相比之下，手写文本生成（HTG）则是一个新兴的、具有巨大潜力的研究方向。然而，由于个体书写风格的变化莫测，这一任务面临诸多挑战。为了生成逼真的手写文本，需要一个大型且多样化的数据集，但这类数据集难以收集并无法轻易获取。 孟加拉语是世界上使用人数第五多的语言。尽管对于英语和阿拉伯语等语言已有若干研究，但对于孟加拉语的手写文本生成却鲜有关注。为填补这一空白，我们提出了一种用于生成孟加拉语手写单词的方法。为此，我们开发并利用了一个由大约500人提供样本的自收集数据集，参与者来自不同的年龄段和性别群体。所有图像都经过预处理以确保一致性和质量。 我们的方法展示了将输入的纯文本转换为多样化的手写输出的能力。我们认为这项工作有助于推进孟加拉语的手写生成技术，并能够支持该领域进一步的研究。
URL
https://arxiv.org/abs/2512.21694
PDF
-
Neural Probe-Based Hallucination Detection for Large Language Models
2025-12-24 05:10:19Shize Liang, Hongzhi WangAbstract
Large language models(LLMs) excel at text generation and knowledge question-answering tasks, but they are prone to generating hallucinated content, severely limiting their application in high-risk domains. Current hallucination detection methods based on uncertainty estimation and external knowledge retrieval suffer from the limitation that they still produce erroneous content at high confidence levels and rely heavily on retrieval efficiency and knowledge coverage. In contrast, probe methods that leverage the model's hidden-layer states offer real-time and lightweight advantages. However, traditional linear probes struggle to capture nonlinear structures in deep semantic this http URL overcome these limitations, we propose a neural network-based framework for token-level hallucination detection. By freezing language model parameters, we employ lightweight MLP probes to perform nonlinear modeling of high-level hidden states. A multi-objective joint loss function is designed to enhance detection stability and semantic disambiguity. Additionally, we establish a layer position-probe performance response model, using Bayesian optimization to automatically search for optimal probe insertion layers and achieve superior training this http URL results on LongFact, HealthBench, and TriviaQA demonstrate that MLP probes significantly outperform state-of-the-art methods in accuracy, recall, and detection capability under low false-positive conditions.
Abstract (translated)
大型语言模型（LLMs）在文本生成和知识问答任务中表现出色，但它们容易产生虚构内容，这严重限制了其在高风险领域的应用。当前基于不确定性估计和外部知识检索的幻觉检测方法存在局限性：即便自信度很高时仍会产生错误内容，并且高度依赖于检索效率和知识覆盖范围。相比之下，利用模型隐藏层状态的探针方法提供了实时性和轻量级的优势。然而，传统的线性探针难以捕捉深层语义中的非线性结构。为克服这些限制，我们提出了一种基于神经网络的框架来进行标记级别的幻觉检测。通过冻结语言模型参数，我们使用轻量级MLP探针对高级隐藏状态进行非线性建模。设计了一个多目标联合损失函数以增强检测稳定性和语义歧义消除能力。此外，建立了层位置-探针性能响应模型，并采用贝叶斯优化自动搜索最佳的探针插入层，从而实现优越的训练效果。在LongFact、HealthBench和TriviaQA上的实验结果表明，在低假阳性条件下，MLP探针对比当前最先进的方法在准确性、召回率和检测能力方面有显著提升。
URL
https://arxiv.org/abs/2512.20949
PDF
-
NVIDIA Nemotron 3: Efficient and Open Intelligence
2025-12-24 00:24:05NVIDIA, , Aaron Blakeman, Aaron Grattafiori, Aarti Basant, Abhibha Gupta, Abhinav Khattar, Adi Renduchintala, Aditya Vavre, Akanksha Shukla, Akhiad Bercovich, Aleksander Ficek, Aleksandr Shaposhnikov, Alex Kondratenko, Alexander Bukharin, Alexandre Milesi, Ali Taghibakhshi, Alisa Liu, Amelia Barton, Ameya Sunil Mahabaleshwarkar, Amir Klein, Amit Zuker, Amnon Geifman, Amy Shen, Anahita Bhiwandiwalla, Andrew Tao, Anjulie Agrusa, Ankur Verma, Ann Guan, Anubhav Mandarwal, Arham Mehta, Ashwath Aithal, Ashwin Poojary, Asif Ahamed, Asit Mishra, Asma Kuriparambil Thekkumpate, Ayush Dattagupta, Banghua Zhu, Bardiya Sadeghi, Barnaby Simkin, Ben Lanir, Benedikt Schifferer, Besmira Nushi, Bilal Kartal, Bita Darvish Rouhani, Boris Ginsburg, Brandon Norick, Brandon Soubasis, Branislav Kisacanin, Brian Yu, Bryan Catanzaro, Carlo del Mundo, Chantal Hwang, Charles Wang, Cheng-Ping Hsieh, Chenghao Zhang, Chenhan Yu, Chetan Mungekar, Chintan Patel, Chris Alexiuk, Christopher Parisien, Collin Neale, Cyril Meurillon, Damon Mosk-Aoyama, Dan Su, Dane Corneil, Daniel Afrimi, Daniel Lo, Daniel Rohrer, Daniel Serebrenik, Daria Gitman, Daria Levy, Darko Stosic, David Mosallanezhad, Deepak Narayanan, Dhruv Nathawani, Dima Rekesh, Dina Yared, Divyanshu Kakwani, Dong Ahn, Duncan Riach, Dusan Stosic, Edgar Minasyan, Edward Lin, Eileen Long, Eileen Peters Long, Elad Segal, Elena Lantz, Ellie Evans, Elliott Ning, Eric Chung, Eric Harper, Eric Tramel, Erick Galinkin, Erik Pounds, Evan Briones, Evelina Bakhturina, Evgeny Tsykunov, Faisal Ladhak, Fay Wang, Fei Jia, Felipe Soares, Feng Chen, Ferenc Galko, Frank Sun, Frankie Siino, Gal Hubara Agam, Ganesh Ajjanagadde, Gantavya Bhatt, Gargi Prasad, George Armstrong, Gerald Shen, Gorkem Batmaz, Grigor Nalbandyan, Haifeng Qian, Harsh Sharma, Hayley Ross, Helen Ngo, Herbert Hum, Herman Sahota, Hexin Wang, Himanshu Soni, Hiren Upadhyay, Huizi Mao, Huy C Nguyen, Huy Q Nguyen, Iain Cunningham, Ido Galil, Ido Shahaf, Igor Gitman, Ilya Loshchilov, Itamar Schen, Itay Levy, Ivan Moshkov, Izik Golan, Izzy Putterman, Jan Kautz, Jane Polak Scowcroft, Jared Casper, Jatin Mitra, Jeffrey Glick, Jenny Chen, Jesse Oliver, Jian Zhang, Jiaqi Zeng, Jie Lou, Jimmy Zhang, Jinhang Choi, Jining Huang, Joey Conway, Joey Guman, John Kamalu, Johnny Greco, Jonathan Cohen, Joseph Jennings, Joyjit Daw, Julien Veron Vialard, Junkeun Yi, Jupinder Parmar, Kai Xu, Kan Zhu, Kari Briski, Katherine Cheung, Katherine Luna, Keith Wyss, Keshav Santhanam, Kevin Shih, Kezhi Kong, Khushi Bhardwaj, Kirthi Shankar, Krishna C. Puvvada, Krzysztof Pawelec, Kumar Anik, Lawrence McAfee, Laya Sleiman, Leon Derczynski, Li Ding, Lizzie Wei, Lucas Liebenwein, Luis Vega, Maanu Grover, Maarten Van Segbroeck, Maer Rodrigues de Melo, Mahdi Nazemi, Makesh Narsimhan Sreedhar, Manoj Kilaru, Maor Ashkenazi, Marc Romeijn, Marcin Chochowski, Mark Cai, Markus Kliegl, Maryam Moosaei, Matt Kulka, Matvei Novikov, Mehrzad Samadi, Melissa Corpuz, Mengru Wang, Meredith Price, Michael Andersch, Michael Boone, Michael Evans, Miguel Martinez, Mikail Khona, Mike Chrzanowski, Minseok Lee, Mohammad Dabbah, Mohammad Shoeybi, Mostofa Patwary, Nabin Mulepati, Najeeb Nabwani, Natalie Hereth, Nave Assaf, Negar Habibi, Neta Zmora, Netanel Haber, Nicola Sessions, Nidhi Bhatia, Nikhil Jukar, Nikki Pope, Nikolai Ludwig, Nima Tajbakhsh, Nir Ailon, Nirmal Juluru, Nishant Sharma, Oleksii Hrinchuk, Oleksii Kuchaiev, Olivier Delalleau, Oluwatobi Olabiyi, Omer Ullman Argov, Omri Puny, Oren Tropp, Ouye Xie, Parth Chadha, Pasha Shamis, Paul Gibbons, Pavlo Molchanov, Pawel Morkisz, Peter Dykas, Peter Jin, Pinky Xu, Piotr Januszewski, Pranav Prashant Thombre, Prasoon Varshney, Pritam Gundecha, Przemek Tredak, Qing Miao, Qiyu Wan, Rabeeh Karimi Mahabadi, Rachit Garg, Ran El-Yaniv, Ran Zilberstein, Rasoul Shafipour, Rich Harang, Rick Izzo, Rima Shahbazyan, Rishabh Garg, Ritika Borkar, Ritu Gala, Riyad Islam, Robert Hesse, Roger Waleffe, Rohit Watve, Roi Koren, Ruoxi Zhang, Russell Hewett, Russell J. Hewett, Ryan Prenger, Ryan Timbrook, Sadegh Mahdavi, Sahil Modi, Samuel Kriman, Sangkug Lim, Sanjay Kariyappa, Sanjeev Satheesh, Saori Kaji, Satish Pasumarthi, Saurav Muralidharan, Sean Narentharen, Sean Narenthiran, Seonmyeong Bak, Sergey Kashirsky, Seth Poulos, Shahar Mor, Shanmugam Ramasamy, Shantanu Acharya, Shaona Ghosh, Sharath Turuvekere Sreenivas, Shelby Thomas, Shiqing Fan, Shreya Gopal, Shrimai Prabhumoye, Shubham Pachori, Shubham Toshniwal, Shuoyang Ding, Siddharth Singh, Simeng Sun, Smita Ithape, Somshubra Majumdar, Soumye Singhal, Stas Sergienko, Stefania Alborghetti, Stephen Ge, Sugam Dipak Devare, Sumeet Kumar Barua, Suseella Panguluri, Suyog Gupta, Sweta Priyadarshi, Syeda Nahida Akter, Tan Bui, Teodor-Dumitru Ene, Terry Kong, Thanh Do, Tijmen Blankevoort, Tim Moon, Tom Balough, Tomer Asida, Tomer Bar Natan, Tomer Ronen, Tugrul Konuk, Twinkle Vashishth, Udi Karpas, Ushnish De, Vahid Noorozi, Vahid Noroozi, Venkat Srinivasan, Venmugil Elango, Victor Cui, Vijay Korthikanti, Vinay Rao, Vitaly Kurin, Vitaly Lavrukhin, Vladimir Anisimov, Wanli Jiang, Wasi Uddin Ahmad, Wei Du, Wei Ping, Wenfei Zhou, Will Jennings, William Zhang, Wojciech Prazuch, Xiaowei Ren, Yashaswi Karnati, Yejin Choi, Yev Meyer, Yi-Fu Wu, Yian Zhang, Yigong Qin, Ying Lin, Yonatan Geifman, Yonggan Fu, Yoshi Subara, Yoshi Suhara, Yubo Gao, Zach Moshe, Zhen Dong, Zhongbo Zhu, Zihan Liu, Zijia Chen, Zijie YanAbstract
We introduce the Nemotron 3 family of models - Nano, Super, and Ultra. These models deliver strong agentic, reasoning, and conversational capabilities. The Nemotron 3 family uses a Mixture-of-Experts hybrid Mamba-Transformer architecture to provide best-in-class throughput and context lengths of up to 1M tokens. Super and Ultra models are trained with NVFP4 and incorporate LatentMoE, a novel approach that improves model quality. The two larger models also include MTP layers for faster text generation. All Nemotron 3 models are post-trained using multi-environment reinforcement learning enabling reasoning, multi-step tool use, and support granular reasoning budget control. Nano, the smallest model, outperforms comparable models in accuracy while remaining extremely cost-efficient for inference. Super is optimized for collaborative agents and high-volume workloads such as IT ticket automation. Ultra, the largest model, provides state-of-the-art accuracy and reasoning performance. Nano is released together with its technical report and this white paper, while Super and Ultra will follow in the coming months. We will openly release the model weights, pre- and post-training software, recipes, and all data for which we hold redistribution rights.
Abstract (translated)
我们介绍了 Nemotron 3 模型系列 - Nano、Super 和 Ultra。这些模型提供了强大的代理、推理和对话能力。Nemotron 3 系列采用混合专家组合的 Mamba-Transformer 架构，提供业界领先的吞吐量，并支持长达100万个令牌（token）的上下文长度。Super 和 Ultra 模型使用 NVFP4 进行训练，并集成了 LatentMoE，这是一种新颖的方法，可提高模型的质量。这两个较大的模型还包含了 MTP 层以加速文本生成过程。所有的 Nemotron 3 模型都经过多环境强化学习的后期训练，使得它们在推理、多步骤工具使用以及细粒度推理预算控制方面具备更强的能力。Nano 是最小的模型，在准确性上超越了同类模型，并且在推理时非常经济高效。Super 模型针对协作代理和如 IT 工单自动化等高负载工作场景进行了优化。Ultra 是最大的模型，提供了业界领先的准确性和推理性能。Nano 模型将与技术报告和白皮书一同发布，而 Super 和 Ultra 将在未来几个月内陆续推出。我们将公开发布所有这些模型的权重、预训练和后期训练软件、配方以及我们拥有再分发权利的所有数据。
URL
https://arxiv.org/abs/2512.20856
PDF
-
SA-DiffuSeq: Addressing Computational and Scalability Challenges in Long-Document Generation with Sparse Attention
2025-12-23 19:35:02Alexandros Christoforos, Chadbourne DavisAbstract
Diffusion based approaches to long form text generation suffer from prohibitive computational cost and memory overhead as sequence length increases. We introduce SA-DiffuSeq, a diffusion framework that integrates sparse attention to fundamentally improve scalability for long document modeling. By selectively allocating attention within the diffusion process, SA-DiffuSeq significantly reduces computational complexity while maintaining semantic coherence and generation quality. A key component of our method is a soft absorbing state tailored to sparse attention dynamics, which stabilizes diffusion trajectories and accelerates sequence reconstruction. This design improves sampling efficiency and enhances precision in long range dependency modeling. Extensive experiments demonstrate that SA-DiffuSeq consistently surpasses state of the art diffusion baselines in both training efficiency and sampling speed, with especially strong gains on extended sequences. These properties make SA-DiffuSeq well suited for demanding long form applications such as scientific writing, large scale code generation, and multi turn long context dialogue. Overall, our results indicate that incorporating structured sparsity into diffusion models is a promising direction for efficient and expressive long text generation.
Abstract (translated)
基于扩散的方法在生成长篇文本时，随着序列长度的增加，会面临高昂的计算成本和内存开销问题。我们提出了一种名为SA-DiffuSeq的新框架，该框架通过整合稀疏注意力机制，从根本上提高了长文档建模的可扩展性。通过在扩散过程中选择性地分配注意力，SA-DiffuSeq显著降低了计算复杂度，并保持了语义连贯性和生成质量。 我们的方法的关键组件是一个专门针对稀疏注意动态特制的软吸收状态，这有助于稳定扩散轨迹并加速序列重构过程。这种设计提高了采样效率，并增强了对长距离依赖关系建模的精确性。 大量的实验表明，SA-DiffuSeq在训练效率和采样速度方面持续超越现有的基于扩散的方法，在处理扩展序列时尤其表现出色。这些特性使得SA-DiffuSeq非常适合需要生成长篇文本的应用场景，例如科学写作、大规模代码生成以及多轮长上下文对话等。 总体而言，我们的结果表明，在扩散模型中引入结构化稀疏性是实现高效且富有表现力的长文本生成的一个有前景的方向。
URL
https://arxiv.org/abs/2512.20724
PDF
-
MoE-DiffuSeq: Enhancing Long-Document Diffusion Models with Sparse Attention and Mixture of Experts
2025-12-23 18:50:54Alexandros Christoforos, Chadbourne DavisAbstract
We present MoE-DiffuSeq, a mixture of experts based framework for enhancing diffusion models in long document generation. Existing diffusion based text generation models, such as DiffuSeq, suffer from high computational cost and memory overhead when applied to extended sequences. To address these challenges, MoE-DiffuSeq integrates sparse attention with a mixture of experts architecture, enabling efficient and scalable long sequence modeling. Our approach introduces a customized sparse attention mechanism designed to reduce computational complexity while preserving text quality and coherence. In addition, we incorporate a soft absorbing state within the diffusion process to accelerate sequence reconstruction and improve generation precision. Extensive experiments demonstrate that MoE-DiffuSeq significantly improves training efficiency and sampling speed compared to existing diffusion models. These advantages are particularly effective for long document scenarios, including scientific article generation, code repository modeling, and long form dialogue generation. Benchmark results further show that MoE-DiffuSeq improves efficiency, speed, accuracy, and expressiveness, advancing the practical applicability of diffusion models for high quality long form text generation.
Abstract (translated)
我们提出了一种基于专家混合框架的MoE-DiffuSeq，旨在通过增强扩散模型来改进长文档生成。现有的基于扩散的文本生成模型（如DiffuSeq）在应用于扩展序列时会面临高昂的计算成本和内存开销问题。为了解决这些问题，MoE-DiffuSeq将稀疏注意力机制与专家混合架构相结合，从而实现高效且可扩展的长序列建模。我们的方法引入了一种定制化的稀疏注意力机制，旨在减少计算复杂度的同时保持文本质量和连贯性。此外，我们在扩散过程中加入了软吸收状态，以加速序列重构并提高生成精度。 广泛的实验表明，与现有的基于扩散的模型相比，MoE-DiffuSeq在训练效率和采样速度方面显著提升。这些优势尤其适用于长文档场景，包括科学文章生成、代码库建模以及长篇对话生成等领域。基准测试结果显示，MoE-DiffuSeq不仅提高了效率和速度，还提升了准确性和表现力，从而推进了扩散模型在高质量长文本生成领域的实际应用性。
URL
https://arxiv.org/abs/2512.20604
PDF
-
UTDesign: A Unified Framework for Stylized Text Editing and Generation in Graphic Design Images
2025-12-23 16:13:55Yiming Zhao, Yuanpeng Gao, Yuxuan Luo, Jiwei Duan, Shisong Lin, Longfei Xiong, Zhouhui LianAbstract
AI-assisted graphic design has emerged as a powerful tool for automating the creation and editing of design elements such as posters, banners, and advertisements. While diffusion-based text-to-image models have demonstrated strong capabilities in visual content generation, their text rendering performance, particularly for small-scale typography and non-Latin scripts, remains limited. In this paper, we propose UTDesign, a unified framework for high-precision stylized text editing and conditional text generation in design images, supporting both English and Chinese scripts. Our framework introduces a novel DiT-based text style transfer model trained from scratch on a synthetic dataset, capable of generating transparent RGBA text foregrounds that preserve the style of reference glyphs. We further extend this model into a conditional text generation framework by training a multi-modal condition encoder on a curated dataset with detailed text annotations, enabling accurate, style-consistent text synthesis conditioned on background images, prompts, and layout specifications. Finally, we integrate our approach into a fully automated text-to-design (T2D) pipeline by incorporating pre-trained text-to-image (T2I) models and an MLLM-based layout planner. Extensive experiments demonstrate that UTDesign achieves state-of-the-art performance among open-source methods in terms of stylistic consistency and text accuracy, and also exhibits unique advantages compared to proprietary commercial approaches. Code and data for this paper are available at this https URL.
Abstract (translated)
AI辅助的图形设计已作为自动化创建和编辑如海报、横幅和广告等设计元素的强大工具兴起。尽管基于扩散模型的文本到图像生成方法在视觉内容生成方面展示了强大的能力，但它们在小规模排版和非拉丁文字渲染上的性能仍然有限。本文中，我们提出了UTDesign框架，这是一个用于高精度风格化文本编辑及条件文本生成的设计图象统一框架，支持英文和中文脚本。我们的框架引入了一个基于DiT的新型文本样式迁移模型，该模型从合成数据集上训练得出，能够生成保持参考字符样式的透明RGBA文本前景。我们进一步通过在精心策划并带有详细文本注释的数据集上训练一个多模态条件编码器，将此模型扩展为一个条件文本生成框架，使基于背景图像、提示和布局规范的准确且风格一致的文字合成成为可能。最后，我们将方法整合到完全自动化的文本到设计（Text-to-Design, T2D）工作流中，通过融合预训练的文本到图像（T2I）模型及MLLM布局规划器实现这一目标。广泛的实验表明，UTDesign在开源方法中达到了风格一致性和文字准确性方面的顶尖水平，并且相比专有的商业方法也展现出独特的优势。本文代码和数据可在提供的链接获取。
URL
https://arxiv.org/abs/2512.20479
PDF
-
SiamGPT: Quality-First Fine-Tuning for Stable Thai Text Generation
2025-12-22 15:00:25Thittipat Pairatsuppawat, Abhibhu Tachaapornchai, Paweekorn Kusolsomboon, Chutikan Chaiwong, Thodsaporn Chay-intr, Kobkrit Viriyayudhakorn, Nongnuch Ketui, Aslan B. WongAbstract
Open-weights large language models remain difficult to deploy for Thai due to unstable generation under complex instructions, despite strong English performance. To mitigate these limitations, We present SiamGPT-32B, an open-weights model based on Qwen3-32B, fine-tuned with a Quality-First strategy emphasizing curated supervision over data scale. The fine-tuning pipeline combines translated high-complexity English instruction data with a Thai-adapted AutoIF framework for instruction and linguistic constraints. Using supervised fine-tuning only, without continual pretraining or corpus expansion, SiamGPT-32B improves instruction adherence, multi-turn robustness, and linguistic stability. Evaluations on the SEA-HELM benchmark show that SiamGPT-32B achieves the strongest overall performance among similar-scale open-weights Thai models, with consistent gains in instruction following, multi-turn dialogue, and natural language understanding.
Abstract (translated)
开源权重的大规模语言模型在泰语部署中仍然面临挑战，因为它们在处理复杂指令时生成的结果不稳定，尽管这些模型在英语上的表现很强。为了缓解这一问题，我们提出了SiamGPT-32B，这是一个基于Qwen3-32B的开源权重模型，并通过强调精心策划监督而非数据规模的Quality-First策略进行了微调。 该微调管道结合了翻译后的高复杂度英文指令数据与适应泰国语言环境的AutoIF框架，用于处理指令和语言约束。仅使用有监督的微调，而没有持续预训练或语料库扩展的情况下，SiamGPT-32B在遵循指令、多轮对话稳定性和语言稳定性方面都有所改进。 在SEA-HELM基准测试中，SiamGPT-32B展示了与其他类似规模的开源权重泰语模型相比最强的整体性能，并且在遵循指令、多轮对话和自然语言理解方面均有持续提升。
URL
https://arxiv.org/abs/2512.19455
PDF
-
Watch Closely: Mitigating Object Hallucinations in Large Vision-Language Models with Disentangled Decoding
2025-12-22 06:20:53Ruiqi Ma, Yu Yan, Chunhong Zhang, Minghao Yin, XinChao Liu, Zhihong Jin, Zheng HuAbstract
Large Vision-Language Models (LVLMs) bridge the gap between visual and linguistic modalities, demonstrating strong potential across a variety of domains. However, despite significant progress, LVLMs still suffer from severe hallucination issues in object recognition tasks. These models often fail to accurately identify certain objects, leading to text generation that appears fluent but does not correspond to the visual content, which can have serious consequences in real-world applications. Recently, several methods have been proposed to alleviate LVLM hallucinations, but most focus solely on reducing hallucinations in the language modality. To mitigate hallucinations in both the language and visual modalities, we introduce Hallucination Disentangled Decoding (HDD) method that requires no training. HDD enhances the original image by segmenting it and selecting images that augment the original, while also utilizing a blank image to eliminate language prior hallucinations in both the original and segmented images. This design not only reduces the model's dependence on language priors but also enhances its visual performance. (Code: this https URL)
Abstract (translated)
大型视觉-语言模型（LVLM）弥合了视觉和语言模态之间的差距，在多个领域展示了强大的潜力。然而，尽管取得了显著进展，这些模型在物体识别任务中仍然存在严重的幻觉问题。这些模型常常无法准确地识别某些对象，导致生成的文本看似流畅但却与视觉内容不符，这可能对实际应用产生严重后果。最近提出了几种缓解LVLM幻觉的方法，但大多数方法仅专注于减少语言模态中的幻觉。为了解决语言和视觉模态中同时存在的幻觉问题，我们引入了一种无需训练即可使用的幻觉分离解码（HDD）方法。HDD通过分割原始图像并选择增强原始图像的图片，并使用空白图来消除原始和分割图像的语言先验幻觉，从而改进了原始图像。这种设计不仅减少了模型对语言先验的依赖，还提升了其视觉性能。（代码链接：此处提供链接）
URL
https://arxiv.org/abs/2512.19070
PDF
-
Phoneme-based speech recognition driven by large language models and sampling marginalization
2025-12-20 14:13:28Te Ma, Nanjie Li, Hao Huang, Zhijian OuAbstract
Recently, the Large Language Model-based Phoneme-to-Grapheme (LLM-P2G) method has shown excellent performance in speech recognition tasks and has become a feasible direction to replace the traditional WFST decoding method. This framework takes into account both recognition accuracy and system scalability through two-stage modeling of phoneme prediction and text generation. However, the existing LLM-P2G adopts the Top-K Marginalized (TKM) training strategy, and its candidate phoneme sequences rely on beam search generation, which has problems such as insufficient path diversity, low training efficiency, and high resource overhead. To this end, this paper proposes a sampling marginalized training strategy (Sampling-K Marginalized, SKM), which replaces beam search with random sampling to generate candidate paths, improving marginalized modeling and training efficiency. Experiments were conducted on Polish and German datasets, and the results showed that SKM further improved the model learning convergence speed and recognition performance while maintaining the complexity of the model. Comparative experiments with a speech recognition method that uses a projector combined with a large language model (SpeechLLM) also show that the SKM-driven LLM-P2G has more advantages in recognition accuracy and structural simplicity. The study verified the practical value and application potential of this method in cross-language speech recognition systems.
Abstract (translated)
最近，基于大型语言模型的音素到字母（LLM-P2G）方法在语音识别任务中表现出色，并已成为取代传统WFST解码方法的一个可行方向。该框架通过音素预测和文本生成的两阶段建模来兼顾识别准确性和系统可扩展性。然而，现有的LLM-P2G采用的是Top-K边际化（TKM）训练策略，其候选音素序列依赖于束搜索生成，存在路径多样性不足、训练效率低以及资源消耗高的问题。 为此，本文提出了一种采样边际化训练策略（Sampling-K Marginalized, SKM），该策略用随机抽样代替束搜索来生成候选路径，从而改进了边际化建模，并提高了训练效率。在波兰语和德语数据集上的实验表明，SKM进一步加快了模型的学习收敛速度并提升了识别性能，同时保持了模型的复杂度不变。与结合投影仪和大型语言模型（SpeechLLM）的语音识别方法进行比较实验的结果也显示，由SKM驱动的LLM-P2G在识别准确性和结构简洁性方面具有更多优势。这项研究验证了该方法在跨语言语音识别系统中的实际价值和应用潜力。
URL
https://arxiv.org/abs/2512.18371
PDF
