Peng Shi

Zhoujun Cheng, Tianbao Xie, Peng Shi et al. · cambridge

Though end-to-end neural approaches have recently been dominating NLP tasks in both performance and ease-of-use, they lack interpretability and robustness. We propose Binder, a training-free neural-symbolic framework that maps the task input to a program, which (1) allows binding a unified API of language model (LM) functionalities to a programming language (e.g., SQL, Python) to extend its grammar coverage and thus tackle more diverse questions, (2) adopts an LM as both the program parser and the underlying model called by the API during execution, and (3) requires only a few in-context exemplar annotations. Specifically, we employ GPT-3 Codex as the LM. In the parsing stage, with only a few in-context exemplars, Codex is able to identify the part of the task input that cannot be answerable by the original programming language, correctly generate API calls to prompt Codex to solve the unanswerable part, and identify where to place the API calls while being compatible with the original grammar. In the execution stage, Codex can perform versatile functionalities (e.g., commonsense QA, information extraction) given proper prompts in the API calls. Binder achieves state-of-the-art results on WikiTableQuestions and TabFact datasets, with explicit output programs that benefit human debugging. Note that previous best systems are all finetuned on tens of thousands of task-specific samples, while Binder only uses dozens of annotations as in-context exemplars without any training. Our code is available at https://github.com/HKUNLP/Binder .

Peng Shi, Rui Zhang, He Bai et al. · apple-ml

In-context learning using large language models has recently shown surprising results for semantic parsing tasks such as Text-to-SQL translation. Prompting GPT-3 or Codex using several examples of question-SQL pairs can produce excellent results, comparable to state-of-the-art finetuning-based models. However, existing work primarily focuses on English datasets, and it is unknown whether large language models can serve as competitive semantic parsers for other languages. To bridge this gap, our work focuses on cross-lingual Text-to-SQL semantic parsing for translating non-English utterances into SQL queries based on an English schema. We consider a zero-shot transfer learning setting with the assumption that we do not have any labeled examples in the target language (but have annotated examples in English). This work introduces the XRICL framework, which learns to retrieve relevant English exemplars for a given query to construct prompts. We also include global translation exemplars for a target language to facilitate the translation process for large language models. To systematically evaluate our model, we construct two new benchmark datasets, XSpider and XKaggle-dbqa, which include questions in Chinese, Vietnamese, Farsi, and Hindi. Our experiments show that XRICL effectively leverages large pre-trained language models to outperform existing baselines. Data and code are publicly available at https://github.com/Impavidity/XRICL.

Dongyu Ru, Lin Qiu, Xiangkun Hu et al. · amazon-science

Despite Retrieval-Augmented Generation (RAG) showing promising capability in leveraging external knowledge, a comprehensive evaluation of RAG systems is still challenging due to the modular nature of RAG, evaluation of long-form responses and reliability of measurements. In this paper, we propose a fine-grained evaluation framework, RAGChecker, that incorporates a suite of diagnostic metrics for both the retrieval and generation modules. Meta evaluation verifies that RAGChecker has significantly better correlations with human judgments than other evaluation metrics. Using RAGChecker, we evaluate 8 RAG systems and conduct an in-depth analysis of their performance, revealing insightful patterns and trade-offs in the design choices of RAG architectures. The metrics of RAGChecker can guide researchers and practitioners in developing more effective RAG systems. This work has been open sourced at https://github.com/amazon-science/RAGChecker.

He Bai, Tong Wang, Alessandro Sordoni et al. · apple-ml

Class-based language models (LMs) have been long devised to address context sparsity in $n$-gram LMs. In this study, we revisit this approach in the context of neural LMs. We hypothesize that class-based prediction leads to an implicit context aggregation for similar words and thus can improve generalization for rare words. We map words that have a common WordNet hypernym to the same class and train large neural LMs by gradually annealing from predicting the class to token prediction during training. Empirically, this curriculum learning strategy consistently improves perplexity over various large, highly-performant state-of-the-art Transformer-based models on two datasets, WikiText-103 and Arxiv. Our analysis shows that the performance improvement is achieved without sacrificing performance on rare words. Finally, we document other attempts that failed to yield empirical gains, and discuss future directions for the adoption of class-based LMs on a larger scale.

Ali Mousavi, Xin Zhan, He Bai et al. · apple-ml

Datasets that pair Knowledge Graphs (KG) and text together (KG-T) can be used to train forward and reverse neural models that generate text from KG and vice versa. However models trained on datasets where KG and text pairs are not equivalent can suffer from more hallucination and poorer recall. In this paper, we verify this empirically by generating datasets with different levels of noise and find that noisier datasets do indeed lead to more hallucination. We argue that the ability of forward and reverse models trained on a dataset to cyclically regenerate source KG or text is a proxy for the equivalence between the KG and the text in the dataset. Using cyclic evaluation we find that manually created WebNLG is much better than automatically created TeKGen and T-REx. Guided by these observations, we construct a new, improved dataset called LAGRANGE using heuristics meant to improve equivalence between KG and text and show the impact of each of the heuristics on cyclic evaluation. We also construct two synthetic datasets using large language models (LLMs), and observe that these are conducive to models that perform significantly well on cyclic generation of text, but less so on cyclic generation of KGs, probably because of a lack of a consistent underlying ontology.

Dongwon Jung, Peng Shi, Yi Zhang et al.

Large reasoning models improve performance by generating extended chain-of-thought (CoT) reasoning, but this behavior becomes inefficient when applied to LLM agents. Current LLM agents often generate verbose textual reasoning at every decision step and allocate reasoning effort nearly uniformly across turns, leading to substantial inefficiency in multi-turn agentic trajectories. We propose Adaptive Latent Agentic Reasoning (ALAR), a dual-mode framework that uses compact latent reasoning for routine turns and selectively escalates to explicit chain-of-thought when deeper deliberation is needed. ALAR learns latent reasoning by using the agent's actions as supervision anchors and is further optimized to use latent reasoning when it is sufficient for task success and reserve explicit CoT for harder decisions. Experiments on agentic search and tool-use benchmarks show that ALAR maintains comparable or better task accuracy while substantially reducing generated tokens by up to 43.6% in search and 84.6% in tool use. These results demonstrate that ALAR improves the accuracy-efficiency trade-off of LLM agents by reducing unnecessary textual reasoning while preserving explicit deliberation for harder decision steps.

Tianbao Xie, Fan Zhou, Zhoujun Cheng et al.

Language agents show potential in being capable of utilizing natural language for varied and intricate tasks in diverse environments, particularly when built upon large language models (LLMs). Current language agent frameworks aim to facilitate the construction of proof-of-concept language agents while neglecting the non-expert user access to agents and paying little attention to application-level designs. We present OpenAgents, an open platform for using and hosting language agents in the wild of everyday life. OpenAgents includes three agents: (1) Data Agent for data analysis with Python/SQL and data tools; (2) Plugins Agent with 200+ daily API tools; (3) Web Agent for autonomous web browsing. OpenAgents enables general users to interact with agent functionalities through a web user interface optimized for swift responses and common failures while offering developers and researchers a seamless deployment experience on local setups, providing a foundation for crafting innovative language agents and facilitating real-world evaluations. We elucidate the challenges and opportunities, aspiring to set a foundation for future research and development of real-world language agents.

Yimu Wang, Peng Shi

While recent progress in video-text retrieval has been advanced by the exploration of better representation learning, in this paper, we present a novel multi-grained sparse learning framework, S3MA, to learn an aligned sparse space shared between the video and the text for video-text retrieval. The shared sparse space is initialized with a finite number of sparse concepts, each of which refers to a number of words. With the text data at hand, we learn and update the shared sparse space in a supervised manner using the proposed similarity and alignment losses. Moreover, to enable multi-grained alignment, we incorporate frame representations for better modeling the video modality and calculating fine-grained and coarse-grained similarities. Benefiting from the learned shared sparse space and multi-grained similarities, extensive experiments on several video-text retrieval benchmarks demonstrate the superiority of S3MA over existing methods. Our code is available at https://github.com/yimuwangcs/Better_Cross_Modal_Retrieval.

Yimu Wang, Peng Shi, Hongyang Zhang

In this paper, we study the problem of generating obstinate (over-stability) adversarial examples by word substitution in NLP, where input text is meaningfully changed but the model's prediction does not, even though it should. Previous word substitution approaches have predominantly focused on manually designed antonym-based strategies for generating obstinate adversarial examples, which hinders its application as these strategies can only find a subset of obstinate adversarial examples and require human efforts. To address this issue, in this paper, we introduce a novel word substitution method named GradObstinate, a gradient-based approach that automatically generates obstinate adversarial examples without any constraints on the search space or the need for manual design principles. To empirically evaluate the efficacy of GradObstinate, we conduct comprehensive experiments on five representative models (Electra, ALBERT, Roberta, DistillBERT, and CLIP) finetuned on four NLP benchmarks (SST-2, MRPC, SNLI, and SQuAD) and a language-grounding benchmark (MSCOCO). Extensive experiments show that our proposed GradObstinate generates more powerful obstinate adversarial examples, exhibiting a higher attack success rate compared to antonym-based methods. Furthermore, to show the transferability of obstinate word substitutions found by GradObstinate, we replace the words in four representative NLP benchmarks with their obstinate substitutions. Notably, obstinate substitutions exhibit a high success rate when transferred to other models in black-box settings, including even GPT-3 and ChatGPT. Examples of obstinate adversarial examples found by GradObstinate are available at https://huggingface.co/spaces/anonauthors/SecretLanguage.

Deyang Jiang, Jing Huang, Xuanle Zhao et al.

Effectively scaling GUI automation is essential for computer-use agents (CUAs); however, existing work primarily focuses on scaling GUI grounding rather than the more crucial GUI planning, which requires more sophisticated data collection. In reality, the exploration process of a CUA across apps/desktops/web pages typically follows a tree structure, with earlier functional entry points often being explored more frequently. Thus, organizing large-scale trajectories into tree structures can reduce data cost and streamline the data scaling of GUI planning. In this work, we propose TreeCUA to efficiently scale GUI automation with tree-structured verifiable evolution. We propose a multi-agent collaborative framework to explore the environment, verify actions, summarize trajectories, and evaluate quality to generate high-quality and scalable GUI trajectories. To improve efficiency, we devise a novel tree-based topology to store and replay duplicate exploration nodes, and design an adaptive exploration algorithm to balance the depth (\emph{i.e.}, trajectory difficulty) and breadth (\emph{i.e.}, trajectory diversity). Moreover, we develop world knowledge guidance and global memory backtracking to avoid low-quality generation. Finally, we naturally extend and propose the TreeCUA-DPO method from abundant tree node information, improving GUI planning capability by referring to the branch information of adjacent trajectories. Experimental results show that TreeCUA and TreeCUA-DPO offer significant improvements, and out-of-domain (OOD) studies further demonstrate strong generalization. All trajectory node information and code will be available at https://github.com/UITron-hub/TreeCUA.

Siqi Yang, Zilve Gao, Haibo Qiu et al.

Multimodal Large Language Models (MLLMs) demonstrate significant potential but remain brittle in complex, long-chain visual reasoning tasks. A critical failure mode is "visual forgetting", where models progressively lose visual grounding as reasoning extends, a phenomenon aptly described as "think longer, see less". We posit this failure stems from current training paradigms prematurely entangling two distinct cognitive skills: (1) abstract logical reasoning "how-to-think") and (2) strategic visual perception ("when-to-look"). This creates a foundational cold-start deficiency -- weakening abstract reasoning -- and a strategic perception deficit, as models lack a policy for when to perceive. In this paper, we propose a novel curriculum-based framework to disentangle these skills. First, we introduce a disentangled Supervised Fine-Tuning (SFT) curriculum that builds a robust abstract reasoning backbone on text-only data before anchoring it to vision with a novel Perception-Grounded Chain-of-Thought (PG-CoT) paradigm. Second, we resolve the strategic perception deficit by formulating timing as a reinforcement learning problem. We design a Pivotal Perception Reward that teaches the model when to look by coupling perceptual actions to linguistic markers of cognitive uncertainty (e.g., "wait", "verify"), thereby learning an autonomous grounding policy. Our contributions include the formalization of these two deficiencies and the development of a principled, two-stage framework to address them, transforming the model from a heuristic-driven observer to a strategic, grounded reasoner. \textbf{Code}: \url{https://github.com/gaozilve-max/learning-when-to-look}.

Sarkar Snigdha Sarathi Das, Ranran Haoran Zhang, Peng Shi et al.

Unified Sequence Labeling that articulates different sequence labeling problems such as Named Entity Recognition, Relation Extraction, Semantic Role Labeling, etc. in a generalized sequence-to-sequence format opens up the opportunity to make the maximum utilization of large language model knowledge toward structured prediction. Unfortunately, this requires formatting them into specialized augmented format unknown to the base pretrained language model (PLMs) necessitating finetuning to the target format. This significantly bounds its usefulness in data-limited settings where finetuning large models cannot properly generalize to the target format. To address this challenge and leverage PLM knowledge effectively, we propose FISH-DIP, a sample-aware dynamic sparse finetuning strategy that selectively focuses on a fraction of parameters, informed by feedback from highly regressing examples, during the fine-tuning process. By leveraging the dynamism of sparsity, our approach mitigates the impact of well-learned samples and prioritizes underperforming instances for improvement in generalization. Across five tasks of sequence labeling, we demonstrate that FISH-DIP can smoothly optimize the model in low resource settings offering upto 40% performance improvements over full fine-tuning depending on target evaluation settings. Also, compared to in-context learning and other parameter-efficient fine-tuning approaches, FISH-DIP performs comparably or better, notably in extreme low-resource settings.

Meituan LongCat Team, Bairui Wang, Bayan et al.

We introduce LongCat-Flash-Omni, a state-of-the-art open-source omni-modal model with 560 billion parameters, excelling at real-time audio-visual interaction. By adopting a curriculum-inspired progressive training strategy that transitions from simpler to increasingly complex modality sequence modeling tasks, LongCat-Flash-Omni attains comprehensive multimodal capabilities while maintaining strong unimodal capability. Building upon LongCat-Flash, which adopts a high-performance Shortcut-connected Mixture-of-Experts (MoE) architecture with zero-computation experts, LongCat-Flash-Omni integrates efficient multimodal perception and speech reconstruction modules. Despite its immense size of 560B parameters (with 27B activated), LongCat-Flash-Omni achieves low-latency real-time audio-visual interaction. For training infrastructure, we developed a modality-decoupled parallelism scheme specifically designed to manage the data and model heterogeneity inherent in large-scale multimodal training. This innovative approach demonstrates exceptional efficiency by sustaining over 90% of the throughput achieved by text-only training. Extensive evaluations show that LongCat-Flash-Omni achieves state-of-the-art performance on omni-modal benchmarks among open-source models. Furthermore, it delivers highly competitive results across a wide range of modality-specific tasks, including text, image, and video understanding, as well as audio understanding and generation. We provide a comprehensive overview of the model architecture design, training procedures, and data strategies, and open-source the model to foster future research and development in the community.

Hideo Kobayashi, Wuwei Lan, Peng Shi et al.

While significant progress has been made on the text-to-SQL task, recent solutions repeatedly encode the same database schema for every question, resulting in unnecessary high inference cost and often overlooking crucial database knowledge. To address these issues, we propose You Only Read Once (YORO), a novel paradigm that directly internalizes database knowledge into the parametric knowledge of a text-to-SQL model during training and eliminates the need for schema encoding during inference. YORO significantly reduces the input token length by 66%-98%. Despite its shorter inputs, our empirical results demonstrate YORO's competitive performances with traditional systems on three benchmarks as well as its significant outperformance on large databases. Furthermore, YORO excels in handling questions with challenging value retrievals such as abbreviation.

Haibo Qiu, Xiaohan Lan, Fanfan Liu et al.

Recent advancements in large language models (LLMs) have witnessed a surge in the development of advanced reasoning paradigms, which are now being integrated into multimodal large language models (MLLMs). However, existing approaches often fall short: methods solely employing reinforcement learning (RL) can struggle with sample inefficiency and activating entirely absent reasoning capabilities, while conventional pipelines that initiate with a cold-start supervised fine-tuning (SFT) phase before RL may restrict the model's exploratory capacity and face suboptimal convergence. In this work, we introduce \textbf{Metis-RISE} (\textbf{R}L \textbf{I}ncentivizes and \textbf{S}FT \textbf{E}nhances) for multimodal reasoning model learning. Unlike conventional approaches, Metis-RISE distinctively omits an initial SFT stage, beginning instead with an RL phase (e.g., using a Group Relative Policy Optimization variant) to incentivize and activate the model's latent reasoning capacity. Subsequently, the targeted SFT stage addresses two key challenges identified during RL: (1) \textit{inefficient trajectory sampling} for tasks where the model possesses but inconsistently applies correct reasoning, which we tackle using self-distilled reasoning trajectories from the RL model itself; and (2) \textit{fundamental capability absence}, which we address by injecting expert-augmented knowledge for prompts where the model entirely fails. This strategic application of RL for incentivization followed by SFT for enhancement forms the core of Metis-RISE, leading to two versions of our MLLMs (7B and 72B parameters). Evaluations on the OpenCompass Multimodal Reasoning Leaderboard demonstrate that both models achieve state-of-the-art performance among similar-sized models, with the 72B version ranking fourth overall. Please refer to our project page for open-source information.

Dongwon Jung, Peng Shi, Yi Zhang

Scaling test-time computation improves large language model performance without additional training. Recent work demonstrates that techniques such as repeated sampling, self-verification, and self-reflection can significantly enhance task success by allocating more inference-time compute. However, applying these techniques across multiple agents in a multi-agent system is difficult: there does not exist principled mechanisms to allocate compute to foster collaboration among agents, to extend test-time scaling to collaborative interactions, or to distribute compute across agents under explicit budget constraints. To address this gap, we propose FutureWeaver, a framework for planning and optimizing test-time compute allocation in multi-agent systems under fixed budgets. FutureWeaver introduces modularized collaboration, formalized as callable functions that encapsulate reusable multi-agent workflows. These modules are automatically derived through self-play reflection by abstracting recurring interaction patterns from past trajectories. Building on these modules, FutureWeaver employs a dual-level planning architecture that optimizes compute allocation by reasoning over the current task state while also speculating on future steps. Experiments on complex agent benchmarks demonstrate that FutureWeaver consistently outperforms baselines across diverse budget settings, validating its effectiveness for multi-agent collaboration in inference-time optimization.

Kun Chen, Peng Shi, Fanfan Liu et al.

Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a critical method for enhancing the reasoning capabilities of Large Language Models (LLMs). However, continuous training often leads to policy entropy collapse, characterized by a rapid decay in entropy that results in premature overconfidence, reduced output diversity, and vanishing gradient norms that inhibit learning. Gradient-Preserving Clipping is a primary factor influencing these dynamics, but existing mitigation strategies are largely static and lack a framework connecting clipping mechanisms to precise entropy control. This paper proposes reshaping entropy control in RL from the perspective of Gradient-Preserving Clipping. We first theoretically and empirically verify the contributions of specific importance sampling ratio regions to entropy growth and reduction. Leveraging these findings, we introduce a novel regulation mechanism using dynamic clipping threshold to precisely manage entropy. Furthermore, we design and evaluate dynamic entropy control strategies, including increase-then-decrease, decrease-increase-decrease, and oscillatory decay. Experimental results demonstrate that these strategies effectively mitigate entropy collapse, and achieve superior performance across multiple benchmarks.

Yilin Cao, Yufeng Zhong, Zhixiong Zeng et al.

Mobile GUI agents have shown strong potential in real-world automation and practical applications. However, most existing agents remain reactive, making decisions mainly from current screen, which limits their performance on long-horizon tasks. Building a world model from repeated interactions enables forecasting action outcomes and supports better decision making for mobile GUI agents. This is challenging because the model must predict post-action states with spatial awareness while remaining efficient enough for practical deployment. In this paper, we propose MobileDreamer, an efficient world-model-based lookahead framework to equip the GUI agents based on the future imagination provided by the world model. It consists of textual sketch world model and rollout imagination for GUI agent. Textual sketch world model forecasts post-action states through a learning process to transform digital images into key task-related sketches, and designs a novel order-invariant learning strategy to preserve the spatial information of GUI elements. The rollout imagination strategy for GUI agent optimizes the action-selection process by leveraging the prediction capability of world model. Experiments on Android World show that MobileDreamer achieves state-of-the-art performance and improves task success by 5.25%. World model evaluations further verify that our textual sketch modeling accurately forecasts key GUI elements.

Fanfan Liu, Youyang Yin, Peng Shi et al.

Recent applications of Reinforcement Learning with Verifiable Rewards (RLVR) to Large Language Models (LLMs) and Vision-Language Models (VLMs) have demonstrated significant success in enhancing reasoning capabilities for complex tasks. During RLVR training, an increase in response length is often regarded as a key factor contributing to the growth of reasoning ability. However, the patterns of change in response length vary significantly across different RLVR algorithms during the training process. To provide a fundamental explanation for these variations, this paper conducts an in-depth analysis of the components of mainstream RLVR algorithms. We present a theoretical analysis of the factors influencing response length and validate our theory through extensive experimentation. Building upon these theoretical findings, we propose the Length-Unbiased Sequence Policy Optimization (LUSPO) algorithm. Specifically, we rectify the length bias inherent in Group Sequence Policy Optimization (GSPO), rendering its loss function unbiased with respect to response length and thereby resolving the issue of response length collapse. We conduct extensive experiments across mathematical reasoning benchmarks and multimodal reasoning scenarios, where LUSPO consistently achieves superior performance. Empirical results demonstrate that LUSPO represents a novel, state-of-the-art optimization strategy compared to existing methods such as GRPO and GSPO.

Meng Wan, Benxi Tian, Jue Wang et al.

The evaluation of time series models has traditionally focused on four canonical tasks: forecasting, imputation, anomaly detection, and classification. While these tasks have driven significant progress, they primarily assess task-specific performance and do not rigorously measure whether a model captures the full generative distribution of the data. We introduce lossless compression as a new paradigm for evaluating time series models, grounded in Shannon's source coding theorem. This perspective establishes a direct equivalence between optimal compression length and the negative log-likelihood, providing a strict and unified information-theoretic criterion for modeling capacity. Then We define a standardized evaluation protocol and metrics. We further propose and open-source a comprehensive evaluation framework TSCom-Bench, which enables the rapid adaptation of time series models as backbones for lossless compression. Experiments across diverse datasets on state-of-the-art models, including TimeXer, iTransformer, and PatchTST, demonstrate that compression reveals distributional weaknesses overlooked by classic benchmarks. These findings position lossless compression as a principled task that complements and extends existing evaluation for time series modeling.

Tianbao Xie, Chen Henry Wu, Peng Shi et al.

Structured knowledge grounding (SKG) leverages structured knowledge to complete user requests, such as semantic parsing over databases and question answering over knowledge bases. Since the inputs and outputs of SKG tasks are heterogeneous, they have been studied separately by different communities, which limits systematic and compatible research on SKG. In this paper, we overcome this limitation by proposing the UnifiedSKG framework, which unifies 21 SKG tasks into a text-to-text format, aiming to promote systematic SKG research, instead of being exclusive to a single task, domain, or dataset. We use UnifiedSKG to benchmark T5 with different sizes and show that T5, with simple modifications when necessary, achieves state-of-the-art performance on almost all of the 21 tasks. We further demonstrate that multi-task prefix-tuning improves the performance on most tasks, largely improving the overall performance. UnifiedSKG also facilitates the investigation of zero-shot and few-shot learning, and we show that T0, GPT-3, and Codex struggle in zero-shot and few-shot learning for SKG. We also use UnifiedSKG to conduct a series of controlled experiments on structured knowledge encoding variants across SKG tasks. UnifiedSKG is easily extensible to more tasks, and it is open-sourced at https://github.com/hkunlp/unifiedskg.

Xinyu Zhang, Xueguang Ma, Peng Shi et al.

We present Mr. TyDi, a multi-lingual benchmark dataset for mono-lingual retrieval in eleven typologically diverse languages, designed to evaluate ranking with learned dense representations. The goal of this resource is to spur research in dense retrieval techniques in non-English languages, motivated by recent observations that existing techniques for representation learning perform poorly when applied to out-of-distribution data. As a starting point, we provide zero-shot baselines for this new dataset based on a multi-lingual adaptation of DPR that we call "mDPR". Experiments show that although the effectiveness of mDPR is much lower than BM25, dense representations nevertheless appear to provide valuable relevance signals, improving BM25 results in sparse-dense hybrids. In addition to analyses of our results, we also discuss future challenges and present a research agenda in multi-lingual dense retrieval. Mr. TyDi can be downloaded at https://github.com/castorini/mr.tydi.

Chang Shu, Yusen Zhang, Xiangyu Dong et al.

Text generation from semantic parses is to generate textual descriptions for formal representation inputs such as logic forms and SQL queries. This is challenging due to two reasons: (1) the complex and intensive inner logic with the data scarcity constraint, (2) the lack of automatic evaluation metrics for logic consistency. To address these two challenges, this paper first proposes SNOWBALL, a framework for logic consistent text generation from semantic parses that employs an iterative training procedure by recursively augmenting the training set with quality control. Second, we propose a novel automatic metric, BLEC, for evaluating the logical consistency between the semantic parses and generated texts. The experimental results on two benchmark datasets, Logic2Text and Spider, demonstrate the SNOWBALL framework enhances the logic consistency on both BLEC and human evaluation. Furthermore, our statistical analysis reveals that BLEC is more logically consistent with human evaluation than general-purpose automatic metrics including BLEU, ROUGE and, BLEURT. Our data and code are available at https://github.com/Ciaranshu/relogic.

Yusen Zhang, Xiangyu Dong, Shuaichen Chang et al.

Neural models have achieved significant results on the text-to-SQL task, in which most current work assumes all the input questions are legal and generates a SQL query for any input. However, in the real scenario, users can input any text that may not be able to be answered by a SQL query. In this work, we propose TriageSQL, the first cross-domain text-to-SQL question intention classification benchmark that requires models to distinguish four types of unanswerable questions from answerable questions. The baseline RoBERTa model achieves a 60% F1 score on the test set, demonstrating the need for further improvement on this task. Our dataset is available at https://github.com/chatc/TriageSQL.

Zongmin Liu, Jirui Wang, Jie Li et al.

To better care for the elderly and disabled, it is essential for service robots to have an effective fusion method of object detection and grasp estimation. However, limited research has been observed on the combination of object detection and grasp estimation. To overcome this technical difficulty, a novel integrated method of detection-grasping for specific object based on the box coordinate matching is proposed in this paper. Firstly, the SOLOv2 instance segmentation model is improved by adding channel attention module (CAM) and spatial attention module (SAM). Then, the atrous spatial pyramid pooling (ASPP) and CAM are added to the generative residual convolutional neural network (GR-CNN) model to optimize grasp estimation. Furthermore, a detection-grasping integrated algorithm based on box coordinate matching (DG-BCM) is proposed to obtain the fusion model of object detection and grasp estimation. For verification, experiments on object detection and grasp estimation are conducted separately to verify the superiority of improved models. Additionally, grasping tasks for several specific objects are implemented on a simulation platform, demonstrating the feasibility and effectiveness of DG-BCM algorithm proposed in this paper.

Harri Vanhems, Yue Zhao, Peng Shi et al.

Standard regression methods typically optimize a single pointwise objective, such as mean squared error, which conflates the learning of ordering with the learning of scale. This coupling renders models vulnerable to outliers and heavy-tailed noise. We propose CAIRO (Calibrate After Initial Rank Ordering), a framework that decouples regression into two distinct stages. In the first stage, we learn a scoring function by minimizing a scale-invariant ranking loss; in the second, we recover the target scale via isotonic regression. We theoretically characterize a class of "Optimal-in-Rank-Order" objectives -- including variants of RankNet and Gini covariance -- and prove that they recover the ordering of the true conditional mean under mild assumptions. We further show that subsequent monotone calibration guarantees recovery of the true regression function. Empirically, CAIRO combines the representation learning of neural networks with the robustness of rank-based statistics. It matches the performance of state-of-the-art tree ensembles on tabular benchmarks and significantly outperforms standard regression objectives in regimes with heavy-tailed or heteroskedastic noise.

Yufeng Zhong, Lei Chen, Zhixiong Zeng et al.

Reading text from images or scanned documents via OCR models has been a longstanding focus of researchers. Intuitively, text reading is perceived as a straightforward perceptual task, and existing work primarily focuses on constructing enriched data engineering to enhance SFT capabilities. In this work, we observe that even advanced OCR models exhibit significantly higher entropy in formatted text (\emph{e.g.}, formula, table, etc.) compared to plain text, often by an order of magnitude. These statistical patterns reveal that advanced OCR models struggle with high output uncertainty when dealing with format sensitive document, suggesting that reasoning over diverse reading pathways may improve OCR performance. To address this, we propose format decoupled reinforcement learning (FD-RL), which leverages high-entropy patterns for targeted optimization. Our approach employs entropy-based data filtration strategy to identify format-intensive instances, and adopt format decoupled rewards tailored to different format types, enabling format-level validation rather than token-level memorization. FD-RL achieves an average score of 90.41 on OmniDocBench, setting a new record for end-to-end models on this highly popular benchmark. More importantly, we conduct comprehensive ablation studies over data, training, filtering, and rewarding strategies, thoroughly validating their effectiveness.

Chi Zhang, Haibo Qiu, Qiming Zhang et al.

Reinforcement Learning with Verifiable Rewards (RLVR) has significantly advanced the reasoning capabilities of Large Language Models (LLMs) and is now being applied to Vision-Language Models (VLMs). However, vanilla RLVR for VLMs verifies only the final textual output, critically neglecting the foundational step of visual perception. This oversight leads to visual hallucinations and reward hacking, as reasoning built upon flawed perception is inherently unreliable. To address this, we propose PEARL (Perceptual-Evidence Anchored Reinforced Learning), a dual-branch, perception-reasoning synergistic that strengthens multimodal reasoning by explicitly anchoring it to verified visual evidence. For each reasoning-oriented QA instance, PEARL first derive a perception checklist -- a set of perception-oriented sub-questions with verifiable answers that probe the model's understanding of key visual evidence. During training, auxiliary rollouts on this checklist yield a perceptual reward that both directly reinforces the model's perception ability and acts as a fidelity gate for reasoning. If the model passes the perception check, its policy update is biased towards evidence-anchored reasoning. Otherwise, the process is halted to prevent reasoning from flawed premises. PEARL can be seamlessly integrated with popular RL methods like GRPO and DAPO. Comprehensive experiments show PEARL achieves substantial gains on multimodal reasoning benchmarks, e.g., a +9.7% improvement over the baseline and +6.6% over GRPO on MathVerse.

Kun Chen, Peng Shi, Haibo Qiu et al.

Reinforcement learning (RL) with verifiable rewards has recently catalyzed a wave of "MLLM-r1" approaches that bring RL to vision language models. Most representative paradigms begin with a cold start, typically employing supervised fine-tuning (SFT), to initialize the policy before RL. However, SFT-based cold start adopts the reasoning paradigm intertwined with task solution and output format, which may induce instruction-style overfitting, weakens out-of-distribution generalization, and ultimately affects downstream RL. We revisit the cold start along two views, its training method and data construction, and introduce the Generalization Factor (GF) coefficient to quantify the generalization capability under different methods. Our empirical study finds that preference-based training methods (e.g. DPO) generalizes better than SFT-based methods in cold start. Motivated by this, we propose SPECS-a Self-distilled, Preference-based Cold Start framework that decouples multimodal learning: (1) generates introspective preference data pairs via self-distillation, avoiding reliance on larger teachers or manual annotation; (2) performs preference-based training to learn, focusing on shallow, transferable surface-form criteria (format, structure, style) rather than memorizing content; and (3) hands off to RL with verifiable rewards for deep reasoning results. Experimental results across multiple multimodal benchmarks show that our decoupling learning framework yields consistent performance gains over strong baselines, improving MEGA-Bench by 4.1% and MathVista by 12.2%. Additional experiments indicate that SPECS contributes to reducing in-distribution "stuckness," improving exploration, stabilizing training, and raising the performance ceiling.

Xiaohan Lan, Fanfan Liu, Haibo Qiu et al.

Inspired by recent advancements in LLM reasoning, the field of multimodal reasoning has seen remarkable progress, achieving significant performance gains on intricate tasks such as mathematical problem-solving. Despite this progress, current multimodal large reasoning models exhibit two key limitations. They tend to employ computationally expensive reasoning even for simple queries, leading to inefficiency. Furthermore, this focus on specialized reasoning often impairs their broader, more general understanding capabilities. In this paper, we propose Metis-HOME: a Hybrid Optimized Mixture-of-Experts framework designed to address this trade-off. Metis-HOME enables a ''Hybrid Thinking'' paradigm by structuring the original dense model into two distinct expert branches: a thinking branch tailored for complex, multi-step reasoning, and a non-thinking branch optimized for rapid, direct inference on tasks like general VQA and OCR. A lightweight, trainable router dynamically allocates queries to the most suitable expert. We instantiate Metis-HOME by adapting the Qwen2.5-VL-7B into an MoE architecture. Comprehensive evaluations reveal that our approach not only substantially enhances complex reasoning abilities but also improves the model's general capabilities, reversing the degradation trend observed in other reasoning-specialized models. Our work establishes a new paradigm for building powerful and versatile MLLMs, effectively resolving the prevalent reasoning-vs-generalization dilemma.

Yusen Zhang, Wenliang Zheng, Aashrith Madasu et al.

High-resolution image (HRI) understanding aims to process images with a large number of pixels, such as pathological images and agricultural aerial images, both of which can exceed 1 million pixels. Vision Large Language Models (VLMs) can allegedly handle HRIs, however, there is a lack of a comprehensive benchmark for VLMs to evaluate HRI understanding. To address this gap, we introduce HRScene, a novel unified benchmark for HRI understanding with rich scenes. HRScene incorporates 25 real-world datasets and 2 synthetic diagnostic datasets with resolutions ranging from 1,024 $\times$ 1,024 to 35,503 $\times$ 26,627. HRScene is collected and re-annotated by 10 graduate-level annotators, covering 25 scenarios, ranging from microscopic to radiology images, street views, long-range pictures, and telescope images. It includes HRIs of real-world objects, scanned documents, and composite multi-image. The two diagnostic evaluation datasets are synthesized by combining the target image with the gold answer and distracting images in different orders, assessing how well models utilize regions in HRI. We conduct extensive experiments involving 28 VLMs, including Gemini 2.0 Flash and GPT-4o. Experiments on HRScene show that current VLMs achieve an average accuracy of around 50% on real-world tasks, revealing significant gaps in HRI understanding. Results on synthetic datasets reveal that VLMs struggle to effectively utilize HRI regions, showing significant Regional Divergence and lost-in-middle, shedding light on future research.

Tianhe Zhang, Suhan Liu, Peng Shi

Fairness has emerged as a critical consideration in the landscape of machine learning algorithms, particularly as AI continues to transform decision-making across societal domains. To ensure that these algorithms are free from bias and do not discriminate against individuals based on sensitive attributes such as gender and race, the field of algorithmic bias has introduced various fairness concepts, along with methodologies to achieve these notions in different contexts. Despite the rapid advancement, not all sectors have embraced these fairness principles to the same extent. One specific sector that merits attention in this regard is insurance. Within the realm of insurance pricing, fairness is defined through a distinct and specialized framework. Consequently, achieving fairness according to established notions does not automatically ensure fair pricing in insurance. In particular, regulators are increasingly emphasizing transparency in pricing algorithms and imposing constraints on insurance companies on the collection and utilization of sensitive consumer attributes. These factors present additional challenges in the implementation of fairness in pricing algorithms. To address these complexities and comply with regulatory demands, we propose an efficient method for constructing fair models that are tailored to the insurance domain, using only privatized sensitive attributes. Notably, our approach ensures statistical guarantees, does not require direct access to sensitive attributes, and adapts to varying transparency requirements, addressing regulatory demands while ensuring fairness in insurance pricing.

Mengyi Gao, Canran Xu, Peng Shi

State-of-the-art approaches to spelling error correction problem include Transformer-based Seq2Seq models, which require large training sets and suffer from slow inference time; and sequence labeling models based on Transformer encoders like BERT, which involve token-level label space and therefore a large pre-defined vocabulary dictionary. In this paper we present a Hierarchical Character Tagger model, or HCTagger, for short text spelling error correction. We use a pre-trained language model at the character level as a text encoder, and then predict character-level edits to transform the original text into its error-free form with a much smaller label space. For decoding, we propose a hierarchical multi-task approach to alleviate the issue of long-tail label distribution without introducing extra model parameters. Experiments on two public misspelling correction datasets demonstrate that HCTagger is an accurate and much faster approach than many existing models.

Naihao Deng, Shuaichen Chang, Peng Shi et al.

Existing text-to-SQL research only considers complete questions as the input, but lay-users might strive to formulate a complete question. To build a smarter natural language interface to database systems (NLIDB) that also processes incomplete questions, we propose a new task, prefix-to-SQL which takes question prefix from users as the input and predicts the intended SQL. We construct a new benchmark called PAGSAS that contains 124K user question prefixes and the intended SQL for 5 sub-tasks Advising, GeoQuery, Scholar, ATIS, and Spider. Additionally, we propose a new metric SAVE to measure how much effort can be saved by users. Experimental results show that PAGSAS is challenging even for strong baseline models such as T5. As we observe the difficulty of prefix-to-SQL is related to the number of omitted tokens, we incorporate curriculum learning of feeding examples with an increasing number of omitted tokens. This improves scores on various sub-tasks by as much as 9% recall scores on sub-task GeoQuery in PAGSAS.

Peng Shi, Rui Zhang, He Bai et al.

Dense retrieval has shown great success in passage ranking in English. However, its effectiveness in document retrieval for non-English languages remains unexplored due to the limitation in training resources. In this work, we explore different transfer techniques for document ranking from English annotations to multiple non-English languages. Our experiments on the test collections in six languages (Chinese, Arabic, French, Hindi, Bengali, Spanish) from diverse language families reveal that zero-shot model-based transfer using mBERT improves the search quality in non-English mono-lingual retrieval. Also, we find that weakly-supervised target language transfer yields competitive performances against the generation-based target language transfer that requires external translators and query generators.

Peng Shi, Tao Yu, Patrick Ng et al.

In this work, we focus on two crucial components in the cross-domain text-to-SQL semantic parsing task: schema linking and value filling. To encourage the model to learn better encoding ability, we propose a column selection auxiliary task to empower the encoder with the relevance matching capability by using explicit learning targets. Furthermore, we propose two value filling methods to build the bridge from the existing zero-shot semantic parsers to real-world applications, considering most of the existing parsers ignore the values filling in the synthesized SQL. With experiments on Spider, our proposed framework improves over the baselines on the execution accuracy and exact set match accuracy when database contents are unavailable, and detailed analysis sheds light on future work.

Peng Shi, Patrick Ng, Zhiguo Wang et al.

Most recently, there has been significant interest in learning contextual representations for various NLP tasks, by leveraging large scale text corpora to train large neural language models with self-supervised learning objectives, such as Masked Language Model (MLM). However, based on a pilot study, we observe three issues of existing general-purpose language models when they are applied to text-to-SQL semantic parsers: fail to detect column mentions in the utterances, fail to infer column mentions from cell values, and fail to compose complex SQL queries. To mitigate these issues, we present a model pre-training framework, Generation-Augmented Pre-training (GAP), that jointly learns representations of natural language utterances and table schemas by leveraging generation models to generate pre-train data. GAP MODEL is trained on 2M utterance-schema pairs and 30K utterance-schema-SQL triples, whose utterances are produced by generative models. Based on experimental results, neural semantic parsers that leverage GAP MODEL as a representation encoder obtain new state-of-the-art results on both SPIDER and CRITERIA-TO-SQL benchmarks.

He Bai, Peng Shi, Jimmy Lin et al.

Transformers are powerful for sequence modeling. Nearly all state-of-the-art language models and pre-trained language models are based on the Transformer architecture. However, it distinguishes sequential tokens only with the token position index. We hypothesize that better contextual representations can be generated from the Transformer with richer positional information. To verify this, we propose a segment-aware Transformer (Segatron), by replacing the original token position encoding with a combined position encoding of paragraph, sentence, and token. We first introduce the segment-aware mechanism to Transformer-XL, which is a popular Transformer-based language model with memory extension and relative position encoding. We find that our method can further improve the Transformer-XL base model and large model, achieving 17.1 perplexity on the WikiText-103 dataset. We further investigate the pre-training masked language modeling task with Segatron. Experimental results show that BERT pre-trained with Segatron (SegaBERT) can outperform BERT with vanilla Transformer on various NLP tasks, and outperforms RoBERTa on zero-shot sentence representation learning.

He Bai, Peng Shi, Jimmy Lin et al.

The semantics of a text is manifested not only by what is read, but also by what is not read. In this article, we will study how the implicit "not read" information such as end-of-paragraph (\eop) and end-of-sequence (\eos) affect the quality of text generation. Specifically, we find that the pre-trained language model GPT2 can generate better continuations by learning to generate the \eop in the fine-tuning stage. Experimental results on English story generation show that \eop can lead to higher BLEU score and lower \eos perplexity. We also conduct experiments on a self-collected Chinese essay dataset with Chinese-GPT2, a character level LM without \eop or \eos during pre-training. Experimental results show that the Chinese GPT2 can generate better essay endings with \eop.

Peng Shi, Jimmy Lin

Recent work has shown the surprising ability of multi-lingual BERT to serve as a zero-shot cross-lingual transfer model for a number of language processing tasks. We combine this finding with a similarly-recently proposal on sentence-level relevance modeling for document retrieval to demonstrate the ability of multi-lingual BERT to transfer models of relevance across languages. Experiments on test collections in five different languages from diverse language families (Chinese, Arabic, French, Hindi, and Bengali) show that models trained with English data improve ranking quality, without any special processing, both for (non-English) mono-lingual retrieval as well as cross-lingual retrieval.

Hsiu-Wei Yang, Yanyan Zou, Peng Shi et al.

Multilingual knowledge graphs (KGs), such as YAGO and DBpedia, represent entities in different languages. The task of cross-lingual entity alignment is to match entities in a source language with their counterparts in target languages. In this work, we investigate embedding-based approaches to encode entities from multilingual KGs into the same vector space, where equivalent entities are close to each other. Specifically, we apply graph convolutional networks (GCNs) to combine multi-aspect information of entities, including topological connections, relations, and attributes of entities, to learn entity embeddings. To exploit the literal descriptions of entities expressed in different languages, we propose two uses of a pretrained multilingual BERT model to bridge cross-lingual gaps. We further propose two strategies to integrate GCN-based and BERT-based modules to boost performance. Extensive experiments on two benchmark datasets demonstrate that our method significantly outperforms existing systems.

Peng Shi, Jimmy Lin

We present simple BERT-based models for relation extraction and semantic role labeling. In recent years, state-of-the-art performance has been achieved using neural models by incorporating lexical and syntactic features such as part-of-speech tags and dependency trees. In this paper, extensive experiments on datasets for these two tasks show that without using any external features, a simple BERT-based model can achieve state-of-the-art performance. To our knowledge, we are the first to successfully apply BERT in this manner. Our models provide strong baselines for future research.

Michael Azmy, Peng Shi, Jimmy Lin et al.

This paper explores the problem of matching entities across different knowledge graphs. Given a query entity in one knowledge graph, we wish to find the corresponding real-world entity in another knowledge graph. We formalize this problem and present two large-scale datasets for this task based on exiting cross-ontology links between DBpedia and Wikidata, focused on several hundred thousand ambiguous entities. Using a classification-based approach, we find that a simple multi-layered perceptron based on representations derived from RDF2Vec graph embeddings of entities in each knowledge graph is sufficient to achieve high accuracy, with only small amounts of training data. The contributions of our work are datasets for examining this problem and strong baselines on which future work can be based.

Peng Shi, Jinfeng Rao, Jimmy Lin

This paper explores the problem of ranking short social media posts with respect to user queries using neural networks. Instead of starting with a complex architecture, we proceed from the bottom up and examine the effectiveness of a simple, word-level Siamese architecture augmented with attention-based mechanisms for capturing semantic "soft" matches between query and post tokens. Extensive experiments on datasets from the TREC Microblog Tracks show that our simple models not only achieve better effectiveness than existing approaches that are far more complex or exploit a more diverse set of relevance signals, but are also much faster. Implementations of our samCNN (Simple Attention-based Matching CNN) models are shared with the community to support future work.

Shuqing Bian, Zhenpeng Deng, Fei Li et al.

Cryptocurrencies (or digital tokens, digital currencies, e.g., BTC, ETH, XRP, NEO) have been rapidly gaining ground in use, value, and understanding among the public, bringing astonishing profits to investors. Unlike other money and banking systems, most digital tokens do not require central authorities. Being decentralized poses significant challenges for credit rating. Most ICOs are currently not subject to government regulations, which makes a reliable credit rating system for ICO projects necessary and urgent. In this paper, we introduce IcoRating, the first learning--based cryptocurrency rating system. We exploit natural-language processing techniques to analyze various aspects of 2,251 digital currencies to date, such as white paper content, founding teams, Github repositories, websites, etc. Supervised learning models are used to correlate the life span and the price change of cryptocurrencies with these features. For the best setting, the proposed system is able to identify scam ICO projects with 0.83 precision. We hope this work will help investors identify scam ICOs and attract more efforts in automatically evaluating and analyzing ICO projects.

Salman Mohammed, Peng Shi, Jimmy Lin

We examine the problem of question answering over knowledge graphs, focusing on simple questions that can be answered by the lookup of a single fact. Adopting a straightforward decomposition of the problem into entity detection, entity linking, relation prediction, and evidence combination, we explore simple yet strong baselines. On the popular SimpleQuestions dataset, we find that basic LSTMs and GRUs plus a few heuristics yield accuracies that approach the state of the art, and techniques that do not use neural networks also perform reasonably well. These results show that gains from sophisticated deep learning techniques proposed in the literature are quite modest and that some previous models exhibit unnecessary complexity.

Yuren Zhou, Jin Wang, Peng Shi et al.

Transportation mode detection with personal devices has been investigated for over ten years due to its importance in monitoring ones' activities, understanding human mobility, and assisting traffic management. However, two main limitations are still preventing it from large-scale deployments: high power consumption, and the lack of high-volume and diverse labeled data. In order to reduce power consumption, existing approaches are sampling using fewer sensors and with lower frequency, which however lead to a lower accuracy. A common way to obtain labeled data is recording the ground truth while collecting data, but such method cannot apply to large-scale deployment due to its inefficiency. To address these issues, we adopt a new low-frequency sampling manner with a hierarchical transportation mode identification algorithm and propose an offline data labeling approach with its manual and automatic implementations. Through a real-world large-scale experiment and comparison with related works, our sampling manner and algorithm are proved to consume much less energy while achieving a competitive accuracy around 85%. The new offline data labeling approach is also validated to be efficient and effective in providing ground truth for model training and testing.