论文检索

Chang Yang, Xinrun Wang, Qinggang Zhang, Qi Jiang, Xiao Huang

World models achieve remarkable success in predicting future states and planning in complex environments and Large Language Models (LLMs) serve as promising foundation to build general world models. However, their performances are usually constrained by the limited external knowledge to specific environments. Existing research attempts to enhance LLM-based world models through prompting or fine-tuning approaches, which are either requiring human knowledge or computationally extensive. Therefore, we introduce Retrieval-Augmented World Models (RAWM), a novel framework that leverages retrieval-augmented generation to efficiently integrate the external knowledge to LLM-based world models. Our main contributions are threefold: (i) We introduce a memory system and design an embedding model to retrieve relevant experiences as the in-context examples to improve the world model’s predictive accuracy. (ii) We develop a reinforcement learning (RL) training pipeline that fine-tunes a small MLP head on the pre-trained embedding model using Proximal Policy Optimization (PPO), further enhancing prediction performance. (iii) We conduct extensive experiments across three diverse environments, i.e., Game24, BlocksWorld, and BabyAI, demonstrating that RAWM consistently outperforms baseline models and exhibits strong generalizability. By leveraging the retrieval-augmented generation and the efficient RL training pipeline, RAWM dynamically utilizes relevant historical experiences and equips LLMs with environment-specific external knowledge without retraining, enabling more accurate and generalizable predictions.

Tianlong Wang, Junzhe Chen, Weibin Liao, Xueting Han, Jing Bai

Reinforcement learning (RL) on self-generated data has emerged as a promising paradigm for improving reasoning in large language models (LLMs). However, RL relies on accurate reward signals, which are scarce in many domains, making it critical to train models that can generalize to unseen problems. Existing methods often focus on task-specific or domain-specific reasoning, lacking consideration for generalization and may degrade performance on other tasks. To address this, we distinguish between abstract plans, representing high-level problem-solving strategies, and concrete solutions, proposing that learning plans develops transferable general reasoning capabilities and promotes better generalization. Building on this insight, we propose PlanLearn, a framework that combines plan-based search with Step-level Advantage Preference Optimization (Step-APO) to optimize plan learning. Experimental results show that PlanLearn, trained exclusively on GSM8K and MATH, not only significantly improves in-domain performance but also enhances out-of-domain benchmarks, such as HumanEval (+12.2%), GPQA (+8.6%), ARC-C (+4.0%), MMLU-STEM (+2.2%), and BBH (+1.8%). The code is available at https://github.com/tianlwang/PlanLearn.

T Karthikeyan, Om Dehlan, Mausam, Manish Gupta

Our goal is to build language model based multi-agent systems for complex planning problems involving multiple explicit and implicit constraints, some of which may be commonsense. Our initial investigations reveal that large language models (LLMs) are often unable to maintain consistency across the planning process, whereas large reasoning models (LRMs) struggle with handling implicit commonsense constraints. In response, we introduce LRPlan, a novel domain-independent, language-based multi-agent architecture where LLM and LRM-based agents collaborate at training time to abstract important patterns, heuristics and insights about the domain. At test time, they collaborate in implementing these learned patterns and insights for a new planning instance. We perform experiments on two datasets, TravelPlanner and TimeArena-Static, and use two LLM-LRM combinations from GPT and DeepSeek families. We find that LRPlan outperforms various multi-agent and single-agent baselines obtaining notably higher accuracy as well as cost efficiency. We make the code publiclyavailable.

Jingsen Zhang, Zihang Tian, Xueyang Feng, Xu Chen, Chong Chen

Generating natural language explanations for recommendations has become increasingly important in recommender systems. Traditional approaches typically treat user reviews as ground truth for explanations and focus on improving review prediction accuracy by designing various model architectures. However, due to limitations in data scale and model capability, these explanations often fail to meet key user-centric aspects such as factuality, personalization, and sentiment coherence, significantly reducing their overall helpfulness to users.In this paper, we propose a novel paradigm that refines initial explanations generated by existing explainable recommender models during the inference stage to enhance their quality in multiple aspects. Specifically, we introduce a multi-agent collaborative refinement framework based on large language models. To ensure alignment between the refinement process and user demands, we employ a plan-then-refine pattern to perform targeted modifications. To enable continuous improvements, we design a hierarchical reflection mechanism that provides feedback to the refinement process from both strategic and content perspectives. Extensive experiments on three datasets demonstrate the effectiveness of our framework.

Liangqi Yuan, Dong-Jun Han, Christopher Brinton, Sabine Brunswicker

The rise of large language models (LLMs) has made natural language-driven route planning an emerging research area that encompasses rich user objectives. Current research exhibits two distinct approaches: direct route planning using LLM-as-Agent and graph-based searching strategies. However, LLMs in the former approach struggle to handle extensive map data, while the latter shows limited capability in understanding natural language preferences. Additionally, a more critical challenge arises from the highly heterogeneous and unpredictable spatio-temporal distribution of users across the globe. In this paper, we introduce a novel LLM-Assisted route Planning (LLMAP) system that employs an LLM-as-Parser to comprehend natural language, identify tasks, and extract user preferences and recognize task dependencies, coupled with a Multi-Step Graph construction with iterative Search (MSGS) algorithm as the underlying solver for optimal route finding. Our multi-objective optimization approach adaptively tunes objective weights to maximize points of interest (POI) quality and task completion rate while minimizing route distance, subject to three key constraints: user time limits, POI opening hours, and task dependencies. We conduct extensive experiments using 1,000 routing prompts sampled with varying complexity across 14 countries and 27 cities worldwide. The results demonstrate that our approach achieves superior performance with guarantees across multiple constraints.

Alex Duchnowski, Ellie Pavlick, Alexander Koller

To investigate the effect of problem presentation on LLMs’ ability to solve optimization problems, we introduce the dataset of Everyday Hard Optimization Problems (EHOP), a collection of NP-hard problems expressed in natural language. EHOP includes problem formulations that could be found in computer science textbooks (e.g., graph coloring), versions that are dressed up as problems that could arise in real life (e.g., party planning), and variants with inverted rules. We find that state-of-the-art LLMs, across multiple prompting strategies, systematically solve textbook problems more accurately than their real-life and inverted counterparts. While reasoning models are more capable, they nonetheless show high variance across problem presentations, suggesting they lack a truly robust reasoning mechanism. We argue that this constitutes evidence that LLMs are still heavily dependent on what was seen in training and struggle to generalize to novel problems.

Zirui Wu, Xiao Liu, Jiayi Li, Lingpeng Kong, Yansong Feng

While Large Language Model-based agents have demonstrated substantial progress in task completion, existing evaluation benchmarks tend to overemphasize single-task performance, with insufficient attention given to the crucial aspects of multitask planning and execution efficiency required in real-world scenarios. To bridge this gap, we present Recipe2Plan, a novel benchmark framework based on real-world cooking scenarios. Unlike conventional benchmarks, Recipe2Plan challenges agents to optimize cooking time through parallel task execution while respecting temporal constraints i.e. specific actions need to be performed within a particular timeintervals following the preceding steps.Overly aggressive local parallelization may disrupt this constraint, potentially compromising the entire cooking process.This strict time constraint between actions raises a unique challenge for agents to balance between maximizing concurrent operations and adhering to critical timing constraints. Extensive experiments with state-of-the-art models reveal challenges in maintaining this balance between efficiency and feasibility. The results highlight the need for improved temporal awareness and global multitasking capabilities in large language models. We open-source our benchmark and code at https://github.com/WilliamZR/Recipe2Plan.

Xiaoran Yin, Xu Luo, Hao Wu, Lianli Gao, Jingkuan Song

The automatic control of mobile devices is essential for efficiently performing complex tasks that involve multiple sequential steps. However, these tasks pose significant challenges due to the limited environmental information available at each step, primarily through visual observations. As a result, current approaches, which typically rely on reactive policies, focus solely on immediate observations and often lead to suboptimal decision-making. To address this problem, we propose Foresighted Planning with World Model-Driven Code Execution (FPWC),a framework that prioritizes natural language understanding and structured reasoning to enhance the agent’s global understanding of the environment by developing a task-oriented, refinable world model at the outset of the task. Foresighted actions are subsequently generated through iterative planning within this world model, executed in the form of executable code. Extensive experiments conducted in simulated environments and on real mobile devices demonstrate that our method outperforms previous approaches, particularly achieving a 44.4% relative improvement in task success rate compared to the state-of-the-art in the simulated environment.

Weimin Xiong, Yifan Song, Qingxiu Dong, Bingchan Zhao, Feifan Song, XWang, Sujian Li

Recent advancements in large language models (LLMs) have enabled LLM-based agents to successfully tackle interactive planning tasks. However, despite their successes, existing approaches often suffer from planning hallucinations and require retraining for each new agent. To address these challenges, we propose the **M**eta **P**lan **O**ptimization (**MPO**) framework, , which enhances agent planning capabilities by directly incorporating explicit guidance. Unlike previous methods that rely on complex knowledge, which either require significant human effort or lack quality assurance, MPO leverages high-level general guidance through meta plans to assist agent planning and enables continuous optimization of the meta plans based on feedback from the agent’s task execution. Our experiments conducted on two representative tasks demonstrate that MPO significantly outperforms existing baselines. Moreover, our analysis indicates that MPO provides a plug-and-play solution that enhances both task completion efficiency and generalization capabilities in previous unseen scenarios.

Dongning Rao, Songlin He, Zhihua Jiang, Ruishi Liang

Currently, large language models (LLMs) based Open domain Natural language planning (LONG) has considerable room for improvement. E.g., non-reusable plans with incomplete intermediate states and missing steps hinder real-world applications. To remedy these flaws, this paper establishes a dataset with a baseline for LONG. The GOLD dataset provides the largest dataset for textual procedures, along with corresponding reusable formal planning domain definitions, to date. The baseline, DIGGER, leverages entity-attribute-level action models, which reveal relevant implicit physical properties (aka attributes) of salient entities in actions. DIGGER first extracts action models and builds typed entity lists from textual procedures. Then, it builds goal states for new tasks and instantiates grounded actions using domain prediction. At last, plans are generalized and translated into textual procedures by LLM. Reference-based metrics, LLM-as-a-Judge, and human evaluation are employed to comprehensively evaluate LONG. Experiments on GOLD validate that DIGGER is stronger and more generalizable than recently proposed approaches and LLMs. I.e., DIGGER is the best in seen domains and applicable to unseen domains without adaptation. Specifically, the BLEU-1 score increased from 0.385 to 0.408 on seen domains and rose to 0.310 on unseen domains.

Lisheng Huang, Yichen Liu, Jinhao Jiang, Rongxiang Zhang, Jiahao Yan, Junyi Li, Wayne Xin Zhao

Recent advances in web-augmented large language models (LLMs) have exhibited strong performance in complex reasoning tasks, yet these capabilities are mostly locked in proprietary systems with opaque architectures. In this work, we propose \textbf{ManuSearch}, a transparent and modular multi-agent framework designed to democratize deep search for LLMs. ManuSearch decomposes the search and reasoning process into three collaborative agents: (1) a solution planning agent that iteratively formulates sub-queries, (2) an Internet search agent that retrieves relevant documents via real-time web search, and (3) a structured webpage reading agent that extracts key evidence from raw web content. To rigorously evaluate deep reasoning abilities, we introduce \textbf{ORION}, a challenging benchmark focused on open-web reasoning over long-tail entities, covering both English and Chinese. Experimental results show that ManuSearch substantially outperforms prior open-source baselines and even surpasses leading closed-source systems. Our work paves the way for reproducible, extensible research in open deep search systems. We release the data and code in [https://github.com/RUCAIBox/ManuSearch](https://github.com/RUCAIBox/ManuSearch).

Van-Thuy Phi, Yuji Matsumoto

This paper presents a unified framework for extracting n-ary property information from materials science literature, addressing the critical challenge of capturing complex relationships that often span multiple sentences. We introduce three complementary approaches: RE-Composition, which transforms binary relations into n-ary structures; Direct EAE, which models polymer properties as events with multiple arguments; and LLM-Guided Assembly, which leverages high-confidence entity and relation outputs to guide structured extraction. Our framework is built upon two novel resources: MatSciNERE, a comprehensive corpus for materials science entities and relations, and PolyEE, a specialized corpus for polymer property events. Through strategic synthetic data generation for both NER and EAE tasks, we achieve significant performance improvements (up to 5.34 F1 points). Experiments demonstrate that our combined approaches outperform any single method, with the LLM-guided approach achieving the highest F1 score (71.53%). The framework enables more comprehensive knowledge extraction from scientific literature, supporting materials discovery and database curation applications. We plan to release our resources and trained models to the research community.

Yu Yan, Sheng Sun, Zhe Wang, Yijun Lin, Zenghao Duan, Zhifei Zheng, Min Liu, Zhiyi Yin, Jianping Zhang

With the development of Large Language Models (LLMs), numerous efforts have revealed their vulnerabilities to jailbreak attacks. Although these studies have driven the progress in LLMs’ safety alignment, it remains unclear whether LLMs have internalized authentic knowledge to deal with real-world crimes, or are merely forced to simulate toxic language patterns. This ambiguity raises concerns that jailbreak success is often attributable to a hallucination loop between jailbroken LLM and judger LLM. By decoupling the use of jailbreak techniques, we construct knowledge-intensive Q&A to investigate the misuse threats of LLMs in terms of dangerous knowledge possession, harmful task planning utility, and harmfulness judgment robustness. Experiments reveal a mismatch between jailbreak success rates and harmful knowledge possession in LLMs, and existing LLM-as-a-judge frameworks tend to anchor harmfulness judgments on toxic language patterns. Our study reveals a gap between existing LLM safety assessments and real-world threat potential.

Cheng Qian, Hongyi Du, Hongru Wang, Xiusi Chen, Yuji Zhang, Avirup Sil, ChengXiang Zhai, Kathleen McKeown, Heng Ji

Recent progress in large language models (LLMs) has enabled substantial advances in solving mathematical problems. However, existing benchmarks often fail to reflect real-world complexity, which demand open-ended, interdisciplinary reasoning and integration of computational tools. To address this gap, we introduce **ModelingBench**, a novel benchmark featuring real-world-inspired, open-ended problems from math modeling competitions across diverse domains, ranging from urban traffic optimization to ecosystem resource planning. These tasks require translating natural language into formal mathematical formulations, applying appropriate tools, and producing structured, defensible reports. ModelingBench supports multiple valid solutions, capturing the ambiguity and creativity of practical modeling. To solve these challenges, we present **ModelingAgent**, a multi-agent framework that coordinates tool use, supports structured workflows, and enables iterative self-refinement to generate well-grounded, creative solutions. Empirical results show that ModelingAgent substantially outperforms strong baselines and often produces solutions indistinguishable from those of human experts. Together, our work provides a comprehensive framework for evaluating and advancing real-world problem-solving in open-ended, interdisciplinary modeling challenges. All the codes are released for future research.

Xiuchao Sui, Daiying Tian, Qi Sun, Ruirui Chen, Dongkyu Choi, Kenneth Kwok, Soujanya Poria

Foundation models (FMs) are increasingly applied to bridge language and action in embodied agents, yet the operational characteristics of different integration strategies remain under-explored—especially for complex instruction following and versatile action generation in changing environments. We investigate three paradigms for robotic systems: end-to-end vision-language-action models (VLAs) that implicitly unify perception and planning, and modular pipelines using either vision-language models (VLMs) or multimodal large language models (MLLMs). Two case studies frame the comparison: instruction grounding, which probs fine-grained language understanding and cross-modal disambiguation; and object manipulation, which targets skill transfer via VLA finetuning. Our experiments reveal trade-offs in system scale, generalization and data efficiency. These findings indicate design lessons for language-driven physical agents and point to challenges and opportunities for FM-powered robotics in real-world conditions.

Minghao Liu, Zhitao He, Zhiyuan Fan, Qingyun Wang, Yi R. Fung

Text-guided image editing has seen significant progress in natural image domains, but its application in medical imaging remains limited and lacks standardized evaluation frameworks. Such editing could revolutionize clinical practices by enabling personalized surgical planning, enhancing medical education, and improving patient communication. To bridge this gap, we introduce MedEBench, a robust benchmark designed to diagnose reliability in text-guided medical image editing. MedEBench consists of 1,182 clinically curated image-prompt pairs covering 70 distinct editing tasks and 13 anatomical regions. It contributes in three key areas: (1) a clinically grounded evaluation framework that measures Editing Accuracy, Context Preservation, and Visual Quality, complemented by detailed descriptions of intended edits and corresponding Region-of-Interest (ROI) masks; (2) a comprehensive comparison of seven state-of-the-art models, revealing consistent patterns of failure; and (3) a diagnostic error analysis technique that leverages attention alignment, using Intersection-over-Union (IoU) between model attention maps and ROI masks to identify mislocalization issues, where models erroneously focus on incorrect anatomical regions. MedEBench sets the stage for developing more reliable and clinically effective text-guided medical image editing tools.

Kinjal Basu, Ibrahim Abdelaziz, Kiran Kate, Mayank Agarwal, Maxwell Crouse, Yara Rizk, Kelsey Bradford, Asim Munawar, Sadhana Kumaravel, Saurabh Goyal 等

The resurgence of autonomous agents built using large language models (LLMs) to solve complex real-world tasks has brought increased focus on LLMs’ fundamental ability of tool or function calling. At the core of these agents, an LLM must plan, execute, and respond using external tools, APIs, and custom functions. Research on tool calling has gathered momentum, but evaluation benchmarks and datasets representing the complexity of the tasks have lagged behind. In this work, we focus on one such complexity, nested sequencing, with the goal of extending existing benchmarks and evaluation. Specifically, we present NESTFUL, a benchmark to evaluate LLMs on nested sequences of API calls, i.e., sequences where the output of one API call is passed as input to a subsequent call. NESTFUL contains 1800+ nested sequences where all the function calls are executable. Experimental results on a variety of models show that the best-performing model (GPT-4o) achieves a full sequence match accuracy of 28% and a win-rate of 60%, necessitating a large scope for improvement in the nested sequencing aspect of function calling. Our analysis of these results provides possible future research directions for the community, in addition to a benchmark to track progress.

Jeonghye Kim, Sojeong Rhee, Minbeom Kim, Dohyung Kim, Sangmook Lee, Youngchul Sung, Kyomin Jung

Recent advances in LLM agents have largely built on reasoning backbones like ReAct, which interleave thought and action in complex environments. However, ReAct often produces ungrounded or incoherent reasoning steps, leading to misalignment between the agent’s actual state and goals. Our analysis finds that this stems from ReAct’s inability to maintain consistent internal beliefs and goal alignment, causing compounding errors and hallucinations. To address this, we introduce ReflAct, a novel backbone that shifts reasoning from merely planning next actions to continuously reflecting on the agent’s state relative to its goal. By explicitly grounding decisions in states and enforcing ongoing goal alignment, ReflAct dramatically improves strategic reliability. This design delivers substantial empirical gains: ReflAct surpasses ReAct by 27.7% on average, achieving a 93.3% success rate in ALFWorld. Notably, ReflAct even outperforms ReAct with added enhancement modules (e.g., Reflexion, WKM), showing that strengthening the core reasoning backbone is key to reliable agent performance.

Yifeng Ding, Hantian Ding, Shiqi Wang, Qing Sun, Varun Kumar, Zijian Wang

Fill-in-the-Middle (FIM), or infilling, has become integral to code language models, enabling generation of missing code given both left and right contexts. However, the current FIM training paradigm which performs next-token prediction (NTP) over reordered sequence often leads to models struggling to generate content that aligns well with the surrounding context. We hypothesize that NTP alone is insufficient for models to learn effective planning conditioned on the distant right context, a critical factor for successful code infilling. To overcome this, we propose Horizon-Length Prediction (HLP), a novel training objective that teaches models to predict the number of remaining middle tokens at each step. HLP advances FIM with lookahead planning, enabling models to inherently learn infilling boundaries for arbitrary left and right contexts without relying on dataset-specific post-processing. Our evaluation across different model families and sizes shows that HLP significantly improves FIM performance by up to 24% relatively on diverse benchmarks, across file-level and repository-level. Furthermore, the enhanced planning capability gained through HLP boosts model performance on code reasoning. Importantly, HLP incurs negligible training overhead and no additional inference cost, ensuring its practicality for real-world scenarios.

Chan Young Park, Jillian Fisher, Marius Memmel, Dipika Khullar, Seoho Yun, Abhishek Gupta, Yejin Choi

Large language models (LLMs) have shown promise in robotic procedural planning, yet their human-centric reasoning often omits the low-level, grounded details needed for robotic execution. Vision-language models (VLMs) offer a path toward more perceptually grounded plans, but current methods either rely on expensive, large-scale models or are constrained to narrow simulation settings. We introduce SelfReVision, a lightweight and scalable self-improvement framework for vision-language procedural planning. SelfReVision enables small VLMs to iteratively critique, revise, and verify their own plans, without external supervision or teacher models, drawing inspiration from chain-of-thought prompting and self-instruct paradigms. Through this self-distillation loop, models generate higher-quality, execution-ready plans that can be used both at inference and for continued fine-tuning. Using models varying from 3B to 72B, our results show that SelfReVision not only boosts performance over weak base VLMs but also outperforms models 100X the size, yielding improved control in downstream embodied tasks.