论文检索

Qinzhuo Wu, Wei Liu, Jian Luan, Bin Wang

Recently, tool-augmented LLMs have gained increasing attention. Given an instruction, tool-augmented LLMs can interact with various external tools in multiple rounds and provide a final answer. However, previous LLMs were trained on overly detailed instructions, which included API names or parameters, while real users would not explicitly mention these API details. This leads to a gap between trained LLMs and real-world scenarios. In addition, most works ignore whether the interaction process follows the instruction. To address these issues, we constructed a training dataset called MGToolBench, which contains statement and category-level instructions to better reflect real-world scenarios. In addition, we propose ToolPlanner, a two-stage reinforcement learning framework that utilizes path planning and two feedback mechanisms to enhance the LLM’s task completion and instruction-following capabilities. Experimental results show that ToolPlanner significantly improves the Match Rate, Pass Rate and Win Rate by 26.8%, 20.2%, and 5.6% compared to the SOTA model. Human evaluation verifies that the multi-granularity instructions can better align with users’ usage habits. Our data and code will be released upon acceptance.

Aseem Srivastava, Smriti Joshi, Tanmoy Chakraborty, Md Shad Akhtar

In mental health counseling, condensing dialogues into concise and relevant summaries (aka counseling notes) holds pivotal significance. Large Language Models (LLMs) exhibit remarkable capabilities in various generative tasks; however, their adaptation to domain-specific intricacies remains challenging, especially within mental health contexts. Unlike standard LLMs, mental health experts first plan to apply domain knowledge in writing summaries. Our work enhances LLMs’ ability by introducing a novel planning engine to orchestrate structuring knowledge alignment. To achieve high-order planning, we divide knowledge encapsulation into two major phases: (i) holding dialogue structure and (ii) incorporating domain-specific knowledge. We employ a planning engine on Llama-2, resulting in a novel framework, PIECE. Our proposed system employs knowledge filtering-cum-scaffolding to encapsulate domain knowledge. Additionally, PIECE leverages sheaf convolution learning to enhance its understanding of the dialogue’s structural nuances. We compare PIECE with 14 baseline methods and observe a significant improvement across ROUGE and Bleurt scores. Further, expert evaluation and analyses validate the generation quality to be effective, sometimes even surpassing the gold standard. We further benchmark PIECE with other LLMs and report improvement, including Llama-2 (+2.72%), Mistral (+2.04%), and Zephyr (+1.59%), to justify the generalizability of the planning engine.

Eldar Kurtic, Amir Moeini, Dan Alistarh

We introduce Mathador-LM, a new benchmark for evaluating the mathematical reasoning on large language models (LLMs), combining ruleset interpretation, planning, and problem-solving. This benchmark is inspired by the Mathador game, where the objective is to reach a target number using basic arithmetic operations on a given set of base numbers, following a simple set of rules. We show that, across leading LLMs, we obtain stable average performance while generating benchmark instances dynamically, following a target difficulty level. Thus, our benchmark alleviates concerns about test-set leakage into training data, an issue that often undermines popular benchmarks. Additionally, we conduct a comprehensive evaluation of both open and closed-source state-of-the-art LLMs on Mathador-LM. Our findings reveal that contemporary models struggle with Mathador-LM, scoring significantly lower than average 3rd graders. This stands in stark contrast to their strong performance on popular mathematical reasoning benchmarks. The implementation of Mathador-LM benchmark is available at https://github.com/IST-DASLab/Mathador-LM.

Chunhui Zhang, Yiren Jian, Zhongyu Ouyang, Soroush Vosoughi

This study explores the inherent limitations of large language models (LLMs) from a scaling perspective, focusing on the upper bounds of their cognitive capabilities. We integrate insights from cognitive science to quantitatively examine how LLMs perform on n-back tasks—a benchmark used to assess working memory, which involves temporarily holding and manipulating information. Our findings reveal that despite the increased model size, LLMs still face significant challenges in holding and processing information effectively, especially under complex task conditions. We also assess various prompting strategies, revealing their diverse impacts on LLM performance. The results highlight the struggle of current LLMs to autonomously discover optimal problem-solving patterns without heavily relying on manually corrected prompts. To move beyond these constraints, fundamental improvements in the planning and search of LLMs are essential for them to reason autonomously. Improving these capabilities will reduce the reliance on external corrections and enable LLMs to become more autonomous in their problem-solving processes.

Weizhou Shen, Chenliang Li, Hongzhan Chen, Ming Yan, Xiaojun Quan, Hehong Chen, Ji Zhang, Fei Huang

Large Language Model (LLM) agents significantly extend the capabilities of standalone LLMs, empowering them to interact with external tools (e.g., APIs, functions) and complete various tasks in a self-directed fashion. The challenge of tool use demands that LLMs not only understand user queries and generate answers accurately but also excel in task planning, tool invocation, and result summarization. While traditional works focus on training a single LLM with all these capabilities, performance limitations become apparent, particularly with smaller models. To overcome these challenges, we propose a novel approach that decomposes the aforementioned capabilities into a planner, caller, and summarizer. Each component is implemented by a single LLM that focuses on a specific capability and collaborates with others to accomplish the task. This modular framework facilitates individual updates and the potential use of smaller LLMs for building each capability. To effectively train this framework, we introduce a two-stage training paradigm. First, we fine-tune a backbone LLM on the entire dataset without discriminating sub-tasks, providing the model with a comprehensive understanding of the task. Second, the fine-tuned LLM is used to instantiate the planner, caller, and summarizer respectively, which are continually fine-tuned on respective sub-tasks. Evaluation across various tool-use benchmarks illustrates that our proposed multi-LLM framework surpasses the traditional single-LLM approach, highlighting its efficacy and advantages in tool learning.

Hongru Wang, Rui Wang, Boyang Xue, Heming Xia, Jingtao Cao, Zeming Liu, Jeff Z. Pan, Kam-Fai Wong

Large Language Models (LLMs) can interact with the real world by connecting with versatile external APIs, resulting in better problem-solving and task automation capabilities. Previous research primarily either focuses on APIs with limited arguments from a single source or overlooks the complex dependency relationship between different APIs. However, it is essential to utilize multiple APIs collaboratively from various sources, especially for complex user instructions. In this paper, we introduce MetaBench, the first benchmark to evaluate LLMs’ ability to plan and execute multiple APIs from various sources in order to complete the user’s task. Specifically, we consider two significant challenges in multiple APIs: 1) graph structures: some APIs can be executed independently while others need to be executed one by one, resulting in graph-like execution order; and 2) permission constraints: which source is authorized to execute the API call. We have experimental results on 9 distinct LLMs; e.g., GPT-4o achieves only a 2.0% success rate at the most complex instruction, revealing that the existing state-of-the-art LLMs still cannot perform well in this situation even with the help of in-context learning and finetuning. Our code and data are publicly available at https://github.com/ruleGreen/AppBench.

Ying Su, Zhan Ling, Haochen Shi, Cheng Jiayang, Yauwai Yim, Yangqiu Song

Large language models(LLMs) have been adopted to process textual task description and accomplish procedural planning in embodied AI tasks because of their powerful reasoning ability. However, there is still lack of study on how vision language models(VLMs) behave when multi-modal task inputs are considered. Counterfactual planning that evaluates the model’s reasoning ability over alternative task situations are also under exploited. In order to evaluate the planning ability of both multi-modal and counterfactual aspects, we propose ActPlan-1K. ActPlan-1K is a multi-modal planning benchmark constructed based on ChatGPT and household activity simulator iGibson2. The benchmark consists of 153 activities and 1,187 instances. Each instance describing one activity has a natural language task description and multiple environment images from the simulator. The gold plan of each instance is action sequences over the objects in provided scenes. Both the correctness and commonsense satisfaction are evaluated on typical VLMs. It turns out that current VLMs are still struggling at generating human-level procedural plans for both normal activities and counterfactual activities. We further provide automatic evaluation metrics by finetuning over BLEURT model to facilitate future research on our benchmark.

Jimin Sun, So Yeon Min, Yingshan Chang, Yonatan Bisk

Tools have become a mainstay of LLMs, allowing them to retrieve knowledge not in their weights, to perform tasks on the web, and even to control robots. However, most ontologies and surveys of tool-use have assumed the core challenge for LLMs is choosing the tool. Instead, we introduce a framework for tools more broadly which guides us to explore a model’s ability to detect “silent” tool errors, and reflect on how to plan. This more directly aligns with the increasingly popular use of models as tools. We provide an initial approach to failure recovery with promising results both on a controlled calculator setting and embodied agent planning.

Yiming Huang, Jianwen Luo, Yan Yu, Yitong Zhang, Fangyu Lei, Yifan Wei, Shizhu He, Lifu Huang, Xiao Liu, Jun Zhao 等

We introduce DA-Code, a code generation benchmark specifically designed to assess LLMs on agent-based data science tasks. This benchmark features three core elements: First, the tasks within DA-Code are inherently challenging, setting them apart from traditional code generation tasks and demanding advanced coding skills in grounding and planning. Second, examples in DA-Code are all based on real and diverse data, covering a wide range of complex data wrangling and analytics tasks. Third, to solve the tasks, the models must utilize complex data science programming languages, including Python and SQL, to perform intricate data processing and derive the answers. We set up the benchmark in a controllable and executable environment that aligns with real-world data analysis scenarios and is scalable. The annotators meticulously designed the evaluation suite to ensure the accuracy and robustness of the evaluation. We developed the DA-Agent baseline. Experiments show that although the baseline performs better than other existing frameworks, using the current best LLMs achieves only 30.5% accuracy, leaving ample room for improvement. We release our benchmark at [link](https://github.com/yiyihum/dabench)

Ameya Godbole, Nicholas Monath, Seungyeon Kim, Ankit Singh Rawat, Andrew McCallum, Manzil Zaheer

In text generation, hallucinations refer to the generation of seemingly coherent text that contradicts established knowledge. One compelling hypothesis is that hallucinations occur when a language model is given a generation task outside its parametric knowledge (due to rarity, recency, domain, etc.). A common strategy to address this limitation is to infuse the language models with retrieval mechanisms, providing the model with relevant knowledge for the task. In this paper, we leverage the planning capabilities of instruction-tuned LLMs and analyze how planning can be used to guide retrieval to further reduce the frequency of hallucinations. We empirically evaluate several variations of our proposed approach on long-form text generation tasks. By improving the coverage of relevant facts, plan-guided retrieval and generation can produce more informative responses while providing a higher rate of attribution to source documents.

Sirui Liang, Baoli Zhang, Jun Zhao, Kang Liu

Recent research indicates that large language models (LLMs) possess a certain degree of script planning capability. However, there is still a lack of focused work on evaluating scripts generated by LLMs. The evaluation of scripts poses challenges due to their logical structure, sequential organization, adherence to commonsense constraints, and open-endedness. In this work, We introduced a novel script evaluation dataset, MCScript, consisting of more than 1,500 script evaluation tasks and steps, and developed an agent-based script evaluation framework, ABSEval, to collaboratively evaluate scripts generated by LLMs. Our experiments demonstrate that ABSEval provides superior accuracy and relevance, aligning closely with human evaluation. We evaluated the script planning capabilities of 15 mainstream LLMs and provided a detailed analysis. Furthermore, we observed phenomena like the key factor influencing the script planning ability of LLM is not parameter size and suggested improvements for evaluating open-ended questions.

Yuxiang Zhang, Jing Chen, Junjie Wang, Yaxin Liu, Cheng Yang, Chufan Shi, Xinyu Zhu, Zihao Lin, Hanwen Wan, Yujiu Yang 等

Tool-augmented large language models (LLMs) are rapidly being integrated into real-world applications. Due to the lack of benchmarks, the community has yet to fully understand the hallucination issues within these models. To address this challenge, we introduce a comprehensive diagnostic benchmark, ToolBH. Specifically, we assess the LLM’s hallucinations through two perspectives: depth and breadth. In terms of depth, we propose a multi-level diagnostic process, including (1) solvability detection, (2) solution planning, and (3) missing-tool analysis. For breadth, we consider three scenarios based on the characteristics of the toolset: missing necessary tools, potential tools, and limited functionality tools. Furthermore, we developed seven tasks and collected 700 evaluation samples through multiple rounds of manual annotation. The results show the significant challenges presented by the ToolBH benchmark. The current advanced models Gemini-1.5-Pro and GPT-4o only achieve total scores of 45.3 and 37.0, respectively, on a scale of 100. In this benchmark, larger model parameters do not guarantee better performance; the training data and response strategies also play crucial roles in tool-enhanced LLM scenarios. Our diagnostic analysis indicates that the primary reason for model errors lies in assessing task solvability. Additionally, open-weight models suffer from performance drops with verbose replies, whereas proprietary models excel with longer reasoning.

Shasha Guo, Lizi Liao, Jing Zhang, Cuiping Li, Hong Chen

Conversational Question Generation (CQG) enhances the interactivity of conversational question-answering systems in fields such as education, customer service, and entertainment. However, traditional CQG, focusing primarily on the immediate context, lacks the conversational foresight necessary to guide conversations toward specified conclusions. This limitation significantly restricts their ability to achieve conclusion-oriented conversational outcomes. In this work, we redefine the CQG task as Conclusion-driven Conversational Question Generation (CCQG) by focusing on proactivity, not merely reacting to the unfolding conversation but actively steering it towards a conclusion-oriented question-answer pair. To address this, we propose a novel approach, called Proactive Conversational Question Planning with self-Refining (PCQPR). Concretely, by integrating a planning algorithm inspired by Monte Carlo Tree Search (MCTS) with the analytical capabilities of large language models (LLMs), PCQPR predicts future conversation turns and continuously refines its questioning strategies. This iterative self-refining mechanism ensures the generation of contextually relevant questions strategically devised to reach a specified outcome. Our extensive evaluations demonstrate that PCQPR significantly surpasses existing CQG methods, marking a paradigm shift towards conclusion-oriented conversational question-answering systems.

Chuyi Shang, Amos You, Sanjay Subramanian, Trevor Darrell, Roei Herzig

Recently, image-based Large Multimodal Models (LMMs) have made significant progress in video question-answering (VideoQA) using a frame-wise approach by leveraging large-scale pretraining in a zero-shot manner. Nevertheless, these models need to be capable of finding relevant information, extracting it, and answering the question simultaneously. Currently, existing methods perform all of these steps in a single pass without being able to adapt if insufficient or incorrect information is collected. To overcome this, we introduce a modular multi-LMM agent framework based on several agents with different roles, instructed by a Planner agent that updates its instructions using shared feedback from the other agents. Specifically, we propose TraveLER, a method that can create a plan to "**Trave**rse” through the video, ask questions about individual frames to "**L**ocate” and store key information, and then "**E**valuate” if there is enough information to answer the question. Finally, if there is not enough information, our method is able to "**R**eplan” based on its collected knowledge. Through extensive experiments, we find that the proposed TraveLER approach improves performance on several VideoQA benchmarks without the need to fine-tune on specific datasets. Our code is available at https://github.com/traveler-framework/TraveLER.

Palak Jain, Livio Baldini Soares, Tom Kwiatkowski

We present RICHES, a novel approach that interleaves retrieval with sequence generation tasks. RICHES offers an alternative to conventional RAG systems by eliminating the need for separate retriever and generator. It retrieves documents by directly decoding their contents, constrained on the corpus. Unifying retrieval with generation allows us to adapt to diverse new tasks via prompting alone. RICHES can work with any Instruction-tuned model, without additional training. It provides attributed evidence, supports multi-hop retrievals and interleaves thoughts to plan on what to retrieve next, all within a single decoding pass of the LLM. We demonstrate the strong performance of RICHES across ODQA tasks including attributed and multi-hop QA.

Patomporn Payoungkhamdee, Peerat Limkonchotiwat, Jinheon Baek, Potsawee Manakul, Can Udomcharoenchaikit, Ekapol Chuangsuwanich, Sarana Nutanong

Reasoning is one crucial capability in Large Language Models (LLMs), allowing them to perform complex tasks such as solving math problems and multi-step planning. While reasoning capability can emerge in larger models, smaller ones usually have to rely on distillation to transfer this capability from a larger model. However, recent efforts to distill reasoning capabilities have focused mainly on English, leaving multilingual distillation underexplored. To address this gap, this paper examines existing English reasoning distillation methods that utilize a variety of positive rationales in multilingual settings and proposes d-CoT-nR, a novel approach that incorporates incorrect rationales as additional guidance. Empirical results from multilingual high-school examinations show that d-CoT-nR significantly surpasses the baseline, improving accuracy in unseen languages and correctness in step-by-step reasoning.

Tianyi Men, Pengfei Cao, Zhuoran Jin, Yubo Chen, Kang Liu, Jun Zhao

Planning, as the core module of agents, is crucial in various fields such as embodied agents, web navigation, and tool using. With the development of large language models (LLMs), some researchers treat large language models as intelligent agents to stimulate and evaluate their planning capabilities. However, the planning mechanism is still unclear. In this work, we focus on exploring the look-ahead planning mechanism in large language models from the perspectives of information flow and internal representations. First, we study how planning is done internally by analyzing the multi-layer perception (MLP) and multi-head self-attention (MHSA) components at the last token. We find that the output of MHSA in the middle layers at the last token can directly decode the decision to some extent. Based on this discovery, we further trace the source of MHSA by information flow, and we reveal that MHSA extracts information from spans of the goal states and recent steps. According to information flow, we continue to study what information is encoded within it. Specifically, we explore whether future decisions have been considered in advance in the representation of flow. We demonstrate that the middle and upper layers encode a few short-term future decisions. Overall, our research analyzes the look-ahead planning mechanisms of LLMs, facilitating future research on LLMs performing planning tasks.

Lin Xu, Zhiyuan Hu, Daquan Zhou, Hongyu Ren, Zhen Dong, Kurt Keutzer, See-Kiong Ng, Jiashi Feng

Large Language Models (LLMs) have significantly advanced natural language processing, demonstrating exceptional reasoning, tool usage, and memory capabilities. As their applications expand into multi-agent environments, there arises a need for a comprehensive evaluation framework that captures LLMs’ reasoning, planning, collaboration, and other social abilities. This work introduces a novel competition-based benchmark framework specifically designed to assess LLMs within multi-agent settings, providing quantitative metrics to evaluate their judgment, reasoning, deception, self-awareness, cooperation, coordination, and rationality.We utilize two social deduction games alongside three game-theory scenarios to create diverse environments.Our frame is fortified with the probabilistic graphic modeling (PGM) method, enhancing the LLMs’ capabilities in navigating complex social and cognitive dimensions. We evaluate seven LLMs, quantitatively highlighting a significant capability gap of over threefold between the strongest, GPT o1, and the weakest, Llama-2-70B. It also confirms that our PGM enhancement boosts the abilities of all selected models by an average of 37%. Our data and code can be found here https://github.com/cathyxl/MAgIC.

Qi Zhao, Haotian Fu, Chen Sun, George Konidaris

Long-horizon decision-making tasks present significant challenges for LLM-based agents due to the need for extensive planning over multiple steps. In this paper, we propose a hierarchical framework that decomposes complex tasks into manageable subgoals, utilizing separate LLMs for subgoal prediction and low-level action generation. To address the challenge of creating training signals for unannotated datasets, we develop a reward model that leverages multimodal environment feedback to automatically generate reward signals. We introduce Environment Preference Optimization (EPO), a novel method that generates preference signals from the environment’s feedback and uses them to train LLM-based agents. Extensive experiments on ALFRED demonstrate the state-of-the-art performance of our framework, achieving first place on the ALFRED public leaderboard and showcasing its potential to improve long-horizon decision-making in diverse environments.

Joakim Edin, Maria Maistro, Lars Maaløe, Lasse Borgholt, Jakob Drachmann Havtorn, Tuukka Ruotsalo

Electronic healthcare records are vital for patient safety as they document conditions, plans, and procedures in both free text and medical codes. Language models have significantly enhanced the processing of such records, streamlining workflows and reducing manual data entry, thereby saving healthcare providers significant resources. However, the black-box nature of these models often leaves healthcare professionals hesitant to trust them. State-of-the-art explainability methods increase model transparency but rely on human-annotated evidence spans, which are costly. In this study, we propose an approach to produce plausible and faithful explanations without needing such annotations. We demonstrate on the automated medical coding task that adversarial robustness training improves explanation plausibility and introduce AttInGrad, a new explanation method superior to previous ones. By combining both contributions in a fully unsupervised setup, we produce explanations of comparable quality, or better, to that of a supervised approach. We release our code and model weights.