Kai Song

Xin Zhang, Yang Cao, Baoxing Wu et al.

Large language models have shown strong performance in natural language generation and downstream reasoning tasks, but they still struggle with logical consistency, factual grounding, and interpretability in complex multi-step reasoning. To address these limitations, this paper proposes SGR, a stepwise reasoning enhancement framework that integrates large language models with external knowledge graphs through query-relevant subgraph generation. Given an input question, SGR first extracts key entities, relations, and constraints to construct a structured schema, then retrieves compact subgraphs from a knowledge graph using schema-guided querying. The generated subgraphs provide explicit relational evidence that guides the language model through step-by-step reasoning. In addition, SGR combines direct Cypher-based reasoning with collaborative reasoning integration, allowing candidate answers from multiple reasoning paths to be validated and aggregated according to both model confidence and graph consistency. Experiments on benchmark datasets including CWQ, WebQSP, GrailQA, and KQA Pro demonstrate that SGR improves reasoning accuracy and Hits@1 performance over standard prompting and several knowledge-enhanced baselines. Ablation studies further show that schema guidance and Neo4j-based retrieval are both crucial to the effectiveness of the framework. These results indicate that dynamically generated external subgraphs can improve the accuracy, robustness, and interpretability of LLM-based reasoning.

Yueen Ma, Dafeng Chi, Jingjing Li et al.

The natural language generation domain has witnessed great success thanks to Transformer models. Although they have achieved state-of-the-art generative quality, they often neglect generative diversity. Prior attempts to tackle this issue suffer from either low model capacity or over-complicated architectures. Some recent methods employ the VAE framework to enhance diversity, but their latent variables fully depend on the input context, restricting exploration of the latent space. In this paper, we introduce VOLTA, a framework that elevates generative diversity by bridging Transformer with VAE via a more effective cross-attention-based connection, departing from conventional embedding concatenation or summation. Additionally, we propose integrating InfoGAN-style latent codes to enable input-independent variability, further diversifying the generation. Moreover, our framework accommodates discrete inputs alongside its existing support for continuous inputs. We perform comprehensive experiments with two types of Transformers on six datasets from three different NLG tasks to show that our approach can significantly improve generative diversity while maintaining generative quality.

Yongjing Yin, Junran Ding, Kai Song et al.

Next-token prediction serves as the dominant component in current neural language models. During the training phase, the model employs teacher forcing, which predicts tokens based on all preceding ground truth tokens. However, this approach has been found to create shortcuts, utilizing the revealed prefix to spuriously fit future tokens, potentially compromising the accuracy of the next-token predictor. In this paper, we introduce Semformer, a novel method of training a Transformer language model that explicitly models the semantic planning of response. Specifically, we incorporate a sequence of planning tokens into the prefix, guiding the planning token representations to predict the latent semantic representations of the response, which are induced by an autoencoder. In a minimal planning task (i.e., graph path-finding), our model exhibits near-perfect performance and effectively mitigates shortcut learning, a feat that standard training methods and baseline models have been unable to accomplish. Furthermore, we pretrain Semformer from scratch with 125M parameters, demonstrating its efficacy through measures of perplexity, in-context learning, and fine-tuning on summarization tasks.

Ziyan Zhang, Chao Wang, Zhuo Chen et al.

Reasoning over knowledge graphs (KGs) with first-order logic (FOL) queries is challenging due to the inherent incompleteness of real-world KGs and the compositional complexity of logical query structures. Most existing methods rely on embedding entities and relations into continuous geometric spaces and answer queries via differentiable set operations. While effective for simple query patterns, these approaches often struggle to generalize to complex queries involving multiple operators, deeper reasoning chains, or heterogeneous KG schemas. We propose ROG (Reasoning Over knowledge Graphs with large language models), an ensemble-style framework that combines query-aware KG neighborhood retrieval with large language model (LLM)-based chain-of-thought reasoning. ROG decomposes complex FOL queries into sequences of simpler sub-queries, retrieves compact, query-relevant subgraphs as contextual evidence, and performs step-by-step logical inference using an LLM, avoiding the need for task-specific embedding optimization. Experiments on standard KG reasoning benchmarks demonstrate that ROG consistently outperforms strong embedding-based baselines in terms of mean reciprocal rank (MRR), with particularly notable gains on high-complexity query types. These results suggest that integrating structured KG retrieval with LLM-driven logical reasoning offers a robust and effective alternative for complex KG reasoning tasks.

Ziyan Zhang, Chao Wang, Zhuo Chen et al.

Answering first-order logic (FOL) queries over incomplete knowledge graphs (KGs) is difficult, especially for complex query structures that compose projection, intersection, union, and negation. We propose ROG, a retrieval-augmented framework that combines query-aware neighborhood retrieval with large language model (LLM) chain-of-thought reasoning. ROG decomposes a multi-operator query into a sequence of single-operator sub-queries and grounds each step in compact, query-relevant neighborhood evidence. Intermediate answer sets are cached and reused across steps, improving consistency on deep reasoning chains. This design reduces compounding errors and yields more robust inference on complex and negation-heavy queries. Overall, ROG provides a practical alternative to embedding-based logical reasoning by replacing learned operators with retrieval-grounded, step-wise inference. Experiments on standard KG reasoning benchmarks show consistent gains over strong embedding-based baselines, with the largest improvements on high-complexity and negation-heavy query types.

Xin Zhang, Yang Cao, Baoxing Wu et al.

Large Language Models (LLMs) have demonstrated strong capabilities across diverse NLP applications, such as translation, text generation, and question answering. Nevertheless, they remain limited in complex settings that demand deep reasoning and logical inference. Since these models are trained on large-scale text corpora, their generation process may still introduce irrelevant, noisy, or factually inconsistent content. To mitigate this problem, we introduce SGR, a stepwise framework that enhances LLM reasoning through external subgraph generation. SGR builds query-specific subgraphs from external knowledge bases and uses their semantic structure to support multi-step inference. By grounding intermediate reasoning steps in structured external knowledge, the framework helps the model concentrate on relevant entities, relations, and supporting evidence. In particular, SGR first constructs a subgraph tailored to the input question. It then guides the model to reason progressively over the generated structure and combines multiple reasoning trajectories to obtain the final prediction. Experimental results across several benchmark datasets show that SGR achieves consistent improvements over competitive baselines, highlighting its value for improving both reasoning accuracy and factual reliability.

Xinghao Wang, Pengyu Wang, Xiaoran Liu et al.

Block-sparse attention is promising for accelerating long-context LLM pre-filling, yet identifying relevant blocks efficiently remains a bottleneck. Existing methods typically employ coarse-grained attention as a proxy for block importance estimation, but often resort to expensive token-level searching or scoring, resulting in significant selection overhead. In this work, we trace the inaccuracy of standard coarse-grained attention via mean pooling to a theoretical root cause: the interaction between mean pooling and Rotary Positional Embeddings (RoPE). We prove that mean pooling acts as a low-pass filter that induces destructive interference in high-frequency dimensions, effectively creating a "blind spot" for local positional information (e.g., slash patterns). To address this, we introduce Prism, a training-free spectral-aware approach that decomposes block selection into high-frequency and low-frequency branches. By applying energy-based temperature calibration, Prism restores the attenuated positional signals directly from pooled representations, enabling block importance estimation using purely block-level operations, thereby improving efficiency. Extensive evaluations confirm that Prism maintains accuracy parity with full attention while delivering up to $\mathbf{5.1\times}$ speedup.

Kai Song, Shaofeng Wang, Ziwei Xie et al.

In the user targeting and expanding of new shows on a video platform, the key point is how their embeddings are generated. It's supposed to be personalized from the perspective of both users and shows. Furthermore, the pursue of both instant (click) and long-time (view time) rewards, and the cold-start problem for new shows bring additional challenges. Such a problem is suitable for processing by heterogeneous graph models, because of the natural graph structure of data. But real-world networks usually have billions of nodes and various types of edges. Few existing methods focus on handling large-scale data and exploiting different types of edges, especially the latter. In this paper, we propose a two-stage audience expansion scheme based on an edge-prompted heterogeneous graph network which can take different double-sided interactions and features into account. In the offline stage, to construct the graph, user IDs and specific side information combinations of the shows are chosen to be the nodes, and click/co-click relations and view time are used to build the edges. Embeddings and clustered user groups are then calculated. When new shows arrive, their embeddings and subsequent matching users can be produced within a consistent space. In the online stage, posterior data including click/view users are employed as seeds to look for similar users. The results on the public datasets and our billion-scale data demonstrate the accuracy and efficiency of our approach.

Xinghao Wang, Pengyu Wang, Dong Zhang et al.

Scaling the context length of large language models (LLMs) offers significant benefits but is computationally expensive. This expense stems primarily from the self-attention mechanism, whose $O(N^2)$ complexity with respect to sequence length presents a major bottleneck for both memory and latency. Fortunately, the attention matrix is often sparse, particularly for long sequences, suggesting an opportunity for optimization. Block-sparse attention has emerged as a promising solution that partitions sequences into blocks and skips computation for a subset of these blocks. However, the effectiveness of this method is highly dependent on the underlying attention patterns, which can lead to sub-optimal block-level sparsity. For instance, important key tokens for queries within a single block may be scattered across numerous other blocks, leading to computational redundancy. In this work, we propose Permuted Block-Sparse Attention (\textbf{PBS-Attn}), a plug-and-play method that leverages the permutation properties of attention to increase block-level sparsity and enhance the computational efficiency of LLM prefilling. We conduct comprehensive experiments on challenging real-world long-context datasets, demonstrating that PBS-Attn consistently outperforms existing block-sparse attention methods in model accuracy and closely matches the full attention baseline. Powered by our custom permuted-FlashAttention kernels, PBS-Attn achieves an end-to-end speedup of up to $2.75\times$ in long-context prefilling, confirming its practical viability. Code available at https://github.com/xinghaow99/pbs-attn

Haomin Zhuang, Yujun Zhou, Taicheng Guo et al.

Reinforcement Learning has demonstrated substantial improvements in the reasoning abilities of Large Language Models (LLMs), exhibiting significant applicability across various domains. Recent research has identified that tokens within LLMs play distinct roles during reasoning tasks, categorizing them into high-entropy reasoning tokens and low-entropy knowledge tokens. Prior approaches have typically focused on restricting updates to indirectly encourage exploration, yet they do not explicitly facilitate exploratory behavior during the token generation stage itself. In this work, we introduce a complementary approach that explicitly promotes exploration during sampling by applying distinct temperature settings for different token types. Specifically, our method employs higher temperatures for reasoning tokens to actively encourage exploration, while retaining lower temperatures for knowledge tokens to maintain factual correctness. Furthermore, we systematically investigate various multi-temperature scheduling strategies and their impacts within reinforcement learning contexts. Empirical evaluations on several reasoning benchmarks demonstrate that our approach significantly enhances the reasoning performance of LLMs. The code is available at https://github.com/zhmzm/Multi_Temperature_Verl.git.

Xin Zhang, Yang Cao, Baoxing Wu et al.

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.

Changtai Zhu, Siyin Wang, Ruijun Feng et al.

Conversational search systems require effective handling of context-dependent queries that often contain ambiguity, omission, and coreference. Conversational Query Reformulation (CQR) addresses this challenge by transforming these queries into self-contained forms suitable for off-the-shelf retrievers. However, existing CQR approaches suffer from two critical constraints: high dependency on costly external supervision from human annotations or large language models, and insufficient alignment between the rewriting model and downstream retrievers. We present ConvSearch-R1, the first self-driven framework that completely eliminates dependency on external rewrite supervision by leveraging reinforcement learning to optimize reformulation directly through retrieval signals. Our novel two-stage approach combines Self-Driven Policy Warm-Up to address the cold-start problem through retrieval-guided self-distillation, followed by Retrieval-Guided Reinforcement Learning with a specially designed rank-incentive reward shaping mechanism that addresses the sparsity issue in conventional retrieval metrics. Extensive experiments on TopiOCQA and QReCC datasets demonstrate that ConvSearch-R1 significantly outperforms previous state-of-the-art methods, achieving over 10% improvement on the challenging TopiOCQA dataset while using smaller 3B parameter models without any external supervision.

Chao Wang, Zhuo Chen, Ziyan Zhang et al.

In this paper, we address the challenge of learning with limited fault data for power transformers. Traditional operation and maintenance tools lack effective predictive capabilities for potential faults. The scarcity of extensive fault data makes it difficult to apply machine learning techniques effectively. To solve this problem, we propose a novel approach that leverages the knowledge graph (KG) technology in combination with gradient boosting decision trees (GBDT). This method is designed to efficiently learn from a small set of high-dimensional data, integrating various factors influencing transformer faults and historical operational data. Our approach enables accurate safe state assessments and fault analyses of power transformers despite the limited fault characteristic data. Experimental results demonstrate that this method outperforms other learning approaches in prediction accuracy, such as artificial neural networks (ANN) and logistic regression (LR). Furthermore, it offers significant improvements in progressiveness, practicality, and potential for widespread application.

Mayank Bansal, Milind Chabbi, Kenneth Bogh et al.

Operating a global, real-time platform at Uber's scale requires infrastructure that is both resilient and cost-efficient. Historically, reliability was ensured through a costly 2x capacity model--each service provisioned to handle global traffic independently across two regions--leaving half the fleet idle. We present Uber's Failover Architecture (UFA), which replaces the uniform 2x model with a differentiated architecture aligned to business criticality. Critical services retain failover guarantees, while non-critical services opportunistically use failover buffer capacity reserved for critical services during steady state. During rare "full-peak" failovers, non-critical services are selectively preempted and rapidly restored, with differentiated Service-Level Agreements (SLAs) using on-demand capacity. Automated safeguards, including dependency analysis and regression gates, ensure critical services continue to function even while non-critical services are unavailable. The quantitative impact is significant: UFA reduces steady-state provisioning from 2x to 1.3x, raising utilization from ~20% to ~30% while sustaining 99.97% availability. To date, UFA has hardened over 4,000 unsafe dependencies, eliminated over one million CPU cores from a baseline of about four million cores.

Chenglong Wang, Yang Gan, Hang Zhou et al.

Recent advances in diffusion language models (DLMs) have presented a promising alternative to traditional autoregressive large language models (LLMs). However, DLMs still lag behind LLMs in reasoning performance, especially as the number of denoising steps decreases. Our analysis reveals that this shortcoming arises primarily from the independent generation of masked tokens across denoising steps, which fails to capture the token correlation. In this paper, we define two types of token correlation: intra-sequence correlation and inter-sequence correlation, and demonstrate that enhancing these correlations improves reasoning performance. To this end, we propose a Multi-Reward Optimization (MRO) approach, which encourages DLMs to consider the token correlation during the denoising process. More specifically, our MRO approach leverages test-time scaling, reject sampling, and reinforcement learning to directly optimize the token correlation with multiple elaborate rewards. Additionally, we introduce group step and importance sampling strategies to mitigate reward variance and enhance sampling efficiency. Through extensive experiments, we demonstrate that MRO not only improves reasoning performance but also achieves significant sampling speedups while maintaining high performance on reasoning benchmarks.

Yongyu Mu, Peinan Feng, Zhiquan Cao et al.

In this study, we reveal an in-context learning (ICL) capability of multilingual large language models (LLMs): by translating the input to several languages, we provide Parallel Input in Multiple Languages (PiM) to LLMs, which significantly enhances their comprehension abilities. To test this capability, we design extensive experiments encompassing 8 typical datasets, 7 languages and 8 state-of-the-art multilingual LLMs. Experimental results show that (1) incorporating more languages help PiM surpass the conventional ICL further; (2) even combining with the translations that are inferior to baseline performance can also help. Moreover, by examining the activated neurons in LLMs, we discover a counterintuitive but interesting phenomenon. Contrary to the common thought that PiM would activate more neurons than monolingual input to leverage knowledge learned from diverse languages, PiM actually inhibits neurons and promotes more precise neuron activation especially when more languages are added. This phenomenon aligns with the neuroscience insight about synaptic pruning, which removes less used neural connections, strengthens remainders, and then enhances brain intelligence.

Yinggui Wang, Yuanqing Huang, Jianshu Li et al.

The utilization of personal sensitive data in training face recognition (FR) models poses significant privacy concerns, as adversaries can employ model inversion attacks (MIA) to infer the original training data. Existing defense methods, such as data augmentation and differential privacy, have been employed to mitigate this issue. However, these methods often fail to strike an optimal balance between privacy and accuracy. To address this limitation, this paper introduces an adaptive hybrid masking algorithm against MIA. Specifically, face images are masked in the frequency domain using an adaptive MixUp strategy. Unlike the traditional MixUp algorithm, which is predominantly used for data augmentation, our modified approach incorporates frequency domain mixing. Previous studies have shown that increasing the number of images mixed in MixUp can enhance privacy preservation but at the expense of reduced face recognition accuracy. To overcome this trade-off, we develop an enhanced adaptive MixUp strategy based on reinforcement learning, which enables us to mix a larger number of images while maintaining satisfactory recognition accuracy. To optimize privacy protection, we propose maximizing the reward function (i.e., the loss function of the FR system) during the training of the strategy network. While the loss function of the FR network is minimized in the phase of training the FR network. The strategy network and the face recognition network can be viewed as antagonistic entities in the training process, ultimately reaching a more balanced trade-off. Experimental results demonstrate that our proposed hybrid masking scheme outperforms existing defense algorithms in terms of privacy preservation and recognition accuracy against MIA.

Shuaicheng Liu, Zehao Zhang, Kai Song et al.

The unprecedented performance achieved by deep convolutional neural networks for image classification is linked primarily to their ability of capturing rich structural features at various layers within networks. Here we design a series of experiments, inspired by children's learning of the arithmetic addition of two integers, to showcase that such deep networks can go beyond the structural features to learn deeper knowledge. In our experiments, a set of images is constructed, each image containing an arithmetic addition $n+m$ in its central area, and several classification networks are then trained over a subset of images, using the sum as the label. Tests on the excluded images show that, as the image set gets larger, the networks have well learnt the law of arithmetic additions so as to build up their autonomous reasoning ability strongly. For instance, networks trained over a small percentage of images can classify a big majority of the remaining images correctly, and many arithmetic additions involving some integers that have never been seen during the training can also be solved correctly by the trained networks.

Kai Song, Yue Zhang, Heng Yu et al.

Leveraging user-provided translation to constrain NMT has practical significance. Existing methods can be classified into two main categories, namely the use of placeholder tags for lexicon words and the use of hard constraints during decoding. Both methods can hurt translation fidelity for various reasons. We investigate a data augmentation method, making code-switched training data by replacing source phrases with their target translations. Our method does not change the MNT model or decoding algorithm, allowing the model to learn lexicon translations by copying source-side target words. Extensive experiments show that our method achieves consistent improvements over existing approaches, improving translation of constrained words without hurting unconstrained words.

Kai Song, Yue Zhang, Min Zhang et al.

Neural machine translation (NMT) suffers a performance deficiency when a limited vocabulary fails to cover the source or target side adequately, which happens frequently when dealing with morphologically rich languages. To address this problem, previous work focused on adjusting translation granularity or expanding the vocabulary size. However, morphological information is relatively under-considered in NMT architectures, which may further improve translation quality. We propose a novel method, which can not only reduce data sparsity but also model morphology through a simple but effective mechanism. By predicting the stem and suffix separately during decoding, our system achieves an improvement of up to 1.98 BLEU compared with previous work on English to Russian translation. Our method is orthogonal to different NMT architectures and stably gains improvements on various domains.