Large Language Model~(LLM) based agents have been increasingly popular in solving complex and dynamic tasks, which requires proper evaluation systems to assess their capabilities. Nevertheless, existing benchmarks usually either focus on single-objective tasks or use overly broad assessing metrics, failing to provide a comprehensive inspection of the actual capabilities of LLM-based agents in complicated decision-making tasks. To address these issues, we introduce DSGBench, a more rigorous evaluation platform for strategic decision-making. Firstly, it incorporates six complex strategic games which serve as ideal testbeds due to their long-term and multi-dimensional decision-making demands and flexibility in customizing tasks of various difficulty levels or multiple targets. Secondly, DSGBench employs a fine-grained evaluation scoring system which examines the decision-making capabilities by looking into the performance in five specific dimensions and offering a comprehensive assessment in a well-designed way. Furthermore, DSGBench also incorporates an automated decision-tracking mechanism which enables in-depth analysis of agent behaviour patterns and the changes in their strategies. We demonstrate the advances of DSGBench by applying it to multiple popular LLM-based agents and our results suggest that DSGBench provides valuable insights in choosing LLM-based agents as well as improving their future development. DSGBench is available at https://github.com/DeciBrain-Group/DSGBench.
Large Multimodal Models (LMMs) have emerged as powerful models capable of understanding various data modalities, including text, images, and videos. LMMs encode both text and visual data into tokens that are then combined and processed by an integrated Large Language Model (LLM). Including visual tokens substantially increases the total token count, often by thousands. The increased input length for LLM significantly raises the complexity of inference, resulting in high latency in LMMs. To address this issue, token pruning methods, which remove part of the visual tokens, are proposed. The existing token pruning methods either require extensive calibration and fine-tuning or rely on suboptimal importance metrics which results in increased redundancy among the retained tokens. In this paper, we first formulate token pruning as Max-Min Diversity Problem (MMDP) where the goal is to select a subset such that the diversity among the selected {tokens} is maximized. Then, we solve the MMDP to obtain the selected subset and prune the rest. The proposed method, DivPrune, reduces redundancy and achieves the highest diversity of the selected tokens. By ensuring high diversity, the selected tokens better represent the original tokens, enabling effective performance even at high pruning ratios without requiring fine-tuning. Extensive experiments with various LMMs show that DivPrune achieves state-of-the-art accuracy over 16 image- and video-language datasets. Additionally, DivPrune reduces both the end-to-end latency and GPU memory usage for the tested models. The code is available $\href{https://github.com/vbdi/divprune}{\text{here}}$.
Face manipulation techniques have achieved significant advances, presenting serious challenges to security and social trust. Recent works demonstrate that leveraging multimodal models can enhance the generalization and interpretability of face forgery detection. However, existing annotation approaches, whether through human labeling or direct Multimodal Large Language Model (MLLM) generation, often suffer from hallucination issues, leading to inaccurate text descriptions, especially for high-quality forgeries. To address this, we propose Face Forgery Text Generator (FFTG), a novel annotation pipeline that generates accurate text descriptions by leveraging forgery masks for initial region and type identification, followed by a comprehensive prompting strategy to guide MLLMs in reducing hallucination. We validate our approach through fine-tuning both CLIP with a three-branch training framework combining unimodal and multimodal objectives, and MLLMs with our structured annotations. Experimental results demonstrate that our method not only achieves more accurate annotations with higher region identification accuracy, but also leads to improvements in model performance across various forgery detection benchmarks. Our Codes are available in https://github.com/skJack/VLFFD.git.
Large language model (LLM) agents typically adopt a step-by-step reasoning framework, in which they interleave the processes of thinking and acting to accomplish the given task. However, this paradigm faces a deep-rooted one-pass issue whereby each generated intermediate thought is plugged into the trajectory regardless of its correctness, which can cause irreversible error propagation. To address the issue, this paper proposes a novel framework called Generator-Assistant Stepwise Rollback (GA-Rollback) to induce better decision-making for LLM agents. Particularly, GA-Rollback utilizes a generator to interact with the environment and an assistant to examine each action produced by the generator, where the assistant triggers a rollback operation upon detection of incorrect actions. Moreover, we introduce two additional strategies tailored for the rollback scenario to further improve its effectiveness. Extensive experiments show that GA-Rollback achieves significant improvements over several strong baselines on three widely used benchmarks. Our analysis further reveals that GA-Rollback can function as a robust plug-and-play module, integrating seamlessly with other methods.
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 time intervals 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.
We assess the vulnerability of multimodal large language models to misleading visualizations - charts that distort the underlying data using techniques such as truncated or inverted axes, leading readers to draw inaccurate conclusions that may support misinformation or conspiracy theories. Our analysis shows that these distortions severely harm multimodal large language models, reducing their question-answering accuracy to the level of the random baseline. To mitigate this vulnerability, we introduce six inference-time methods to improve performance of MLLMs on misleading visualizations while preserving their accuracy on non-misleading ones. The most effective approach involves (1) extracting the underlying data table and (2) using a text-only large language model to answer questions based on the table. This method improves performance on misleading visualizations by 15.4 to 19.6 percentage points.
As the use of large language model (LLM) agents continues to grow, their safety vulnerabilities have become increasingly evident. Extensive benchmarks evaluate various aspects of LLM safety by defining the safety relying heavily on general standards, overlooking user-specific standards. However, safety standards for LLM may vary based on a user-specific profiles rather than being universally consistent across all users. This raises a critical research question: Do LLM agents act safely when considering user-specific safety standards? Despite its importance for safe LLM use, no benchmark datasets currently exist to evaluate the user-specific safety of LLMs. To address this gap, we introduce U-SAFEBENCH, the first benchmark designed to assess user-specific aspect of LLM safety. Our evaluation of 18 widely used LLMs reveals current LLMs fail to act safely when considering user-specific safety standards, marking a new discovery in this field. To address this vulnerability, we propose a simple remedy based on chain-of-thought, demonstrating its effectiveness in improving user-specific safety. Our benchmark and code are available at https://github.com/yeonjun-in/U-SafeBench.
Recent advancements in Large Language Model(LLM)-based Multi-Agent Systems(MAS) have demonstrated remarkable potential for tackling complex decision-making tasks. However, existing frameworks inevitably rely on serialized execution paradigms, where agents must complete sequential LLM planning before taking action. This fundamental constraint severely limits real-time responsiveness and adaptation, which is crucial in dynamic environments with ever-changing scenarios. In this paper, we propose a novel parallelized planning-acting framework for LLM-based MAS, featuring a dual-thread architecture with interruptible execution to enable concurrent planning and acting. Specifically, our framework comprises two core threads:(1) a planning thread driven by a centralized memory system, maintaining synchronization of environmental states and agent communication to support dynamic decision-making; and (2) an acting thread equipped with a comprehensive skill library, enabling automated task execution through recursive decomposition. Extensive experiments on challenging Minecraft demonstrate the effectiveness of the proposed framework.
Simultaneous translation of unbounded streaming speech remains a challenging problem due to the need for effectively processing the history speech context and past translations so that quality and latency, including computation overhead, can be balanced. Most prior works assume pre-segmented speech, limiting their real-world applicability. In this paper, we propose InfiniSST, a novel approach that formulates SST as a multi-turn dialogue task, enabling seamless translation of unbounded speech. We construct translation trajectories and robust segments from MuST-C with multi-latency augmentation during training and develop a key-value (KV) cache management strategy to facilitate efficient inference. Experiments on MuST-C En-Es, En-De, and En-Zh demonstrate that InfiniSST reduces computation-aware latency by 0.5 to 1 second while maintaining the same translation quality compared to baselines. Ablation studies further validate the contributions of our data construction and cache management strategy. We release the code at https://github.com/LeiLiLab/InfiniSST
Predicting clinical outcomes from preclinical data is essential for identifying safe and effective drug combinations. Current models rely on structural or target-based features to identify high-efficacy, low-toxicity drug combinations. However, these approaches fail to incorporate the multimodal data necessary for accurate, clinically-relevant predictions. Here, we introduce MADRIGAL, a multimodal AI model that learns from structural, pathway, cell viability, and transcriptomic data to predict drug combination effects across 953 clinical outcomes and 21842 compounds, including combinations of approved drugs and novel compounds in development. MADRIGAL uses a transformer bottleneck module to unify preclinical drug data modalities while handling missing data during training and inference--a major challenge in multimodal learning. It outperforms single-modality methods and state-of-the-art models in predicting adverse drug interactions. MADRIGAL performs virtual screening of anticancer drug combinations and supports polypharmacy management for type II diabetes and metabolic dysfunction-associated steatohepatitis (MASH). It identifies transporter-mediated drug interactions. MADRIGAL predicts resmetirom, the first and only FDA-approved drug for MASH, among therapies with the most favorable safety profile. It supports personalized cancer therapy by integrating genomic profiles from cancer patients. Using primary acute myeloid leukemia samples and patient-derived xenograft models, it predicts the efficacy of personalized drug combinations. Integrating MADRIGAL with a large language model allows users to describe clinical outcomes in natural language, improving safety assessment by identifying potential adverse interactions and toxicity risks. MADRIGAL provides a multimodal approach for designing combination therapies with improved predictive accuracy and clinical relevance.
Molecular optimization is a crucial yet complex and time-intensive process that often acts as a bottleneck for drug development. Traditional methods rely heavily on trial and error, making multi-objective optimization both time-consuming and resource-intensive. Current AI-based methods have shown limited success in handling multi-objective optimization tasks, hampering their practical utilization. To address this challenge, we present MultiMol, a collaborative large language model (LLM) system designed to guide multi-objective molecular optimization. MultiMol comprises two agents, including a data-driven worker agent and a literature-guided research agent. The data-driven worker agent is a large language model being fine-tuned to learn how to generate optimized molecules considering multiple objectives, while the literature-guided research agent is responsible for searching task-related literature to find useful prior knowledge that facilitates identifying the most promising optimized candidates. In evaluations across six multi-objective optimization tasks, MultiMol significantly outperforms existing methods, achieving a 82.30% success rate, in sharp contrast to the 27.50% success rate of current strongest methods. To further validate its practical impact, we tested MultiMol on two real-world challenges. First, we enhanced the selectivity of Xanthine Amine Congener (XAC), a promiscuous ligand that binds both A1R and A2AR, successfully biasing it towards A1R. Second, we improved the bioavailability of Saquinavir, an HIV-1 protease inhibitor with known bioavailability limitations. Overall, these results indicate that MultiMol represents a highly promising approach for multi-objective molecular optimization, holding great potential to accelerate the drug development process and contribute to the advancement of pharmaceutical research.
Reading assessments are essential for enhancing students' comprehension, yet many EdTech applications focus mainly on outcome-based metrics, providing limited insights into student behavior and cognition. This study investigates the use of multimodal data sources -- including eye-tracking data, learning outcomes, assessment content, and teaching standards -- to derive meaningful reading insights. We employ unsupervised learning techniques to identify distinct reading behavior patterns, and then a large language model (LLM) synthesizes the derived information into actionable reports for educators, streamlining the interpretation process. LLM experts and human educators evaluate these reports for clarity, accuracy, relevance, and pedagogical usefulness. Our findings indicate that LLMs can effectively function as educational analysts, turning diverse data into teacher-friendly insights that are well-received by educators. While promising for automating insight generation, human oversight remains crucial to ensure reliability and fairness. This research advances human-centered AI in education, connecting data-driven analytics with practical classroom applications.
Large Language Model (LLM) agents equipped with external tools have become increasingly powerful for complex tasks such as web shopping, automated email replies, and financial trading. However, these advancements amplify the risks of adversarial attacks, especially when agents can access sensitive external functionalities. Nevertheless, manipulating LLM agents into performing targeted malicious actions or invoking specific tools remains challenging, as these agents extensively reason or plan before executing final actions. In this work, we present UDora, a unified red teaming framework designed for LLM agents that dynamically hijacks the agent's reasoning processes to compel malicious behavior. Specifically, UDora first generates the model's reasoning trace for the given task, then automatically identifies optimal points within this trace to insert targeted perturbations. The resulting perturbed reasoning is then used as a surrogate response for optimization. By iteratively applying this process, the LLM agent will then be induced to undertake designated malicious actions or to invoke specific malicious tools. Our approach demonstrates superior effectiveness compared to existing methods across three LLM agent datasets. The code is available at https://github.com/AI-secure/UDora.
Evolutionary Algorithms (EAs) employ random or simplistic selection methods, limiting their exploration of solution spaces and convergence to optimal solutions. The randomness in performing crossover or mutations may limit the model's ability to evolve efficiently. This paper introduces Preference-Aligned Individual Reciprocity (PAIR), a novel selection approach leveraging Large Language Models to emulate human-like mate selection, thereby introducing intelligence to the pairing process in EAs. PAIR prompts an LLM to evaluate individuals within a population based on genetic diversity, fitness level, and crossover compatibility, guiding more informed pairing decisions. We evaluated PAIR against a baseline method called LLM-driven EA (LMEA), published recently. Results indicate that PAIR significantly outperforms LMEA across various TSP instances, achieving lower optimality gaps and improved convergence. This performance is especially noticeable when combined with the flash thinking model, demonstrating increased population diversity to escape local optima. In general, PAIR provides a new strategy in the area of in-context learning for LLM-driven selection in EAs via sophisticated preference modelling, paving the way for improved solutions and further studies into LLM-guided optimization.
Large Language Model (LLM) agents exhibit remarkable performance across diverse applications by using external tools to interact with environments. However, integrating external tools introduces security risks, such as indirect prompt injection (IPI) attacks. Despite defenses designed for IPI attacks, their robustness remains questionable due to insufficient testing against adaptive attacks. In this paper, we evaluate eight different defenses and bypass all of them using adaptive attacks, consistently achieving an attack success rate of over 50%. This reveals critical vulnerabilities in current defenses. Our research underscores the need for adaptive attack evaluation when designing defenses to ensure robustness and reliability. The code is available at https://github.com/uiuc-kang-lab/AdaptiveAttackAgent.
Reducing radiation doses benefits patients, however, the resultant low-dose computed tomography (LDCT) images often suffer from clinically unacceptable noise and artifacts. While deep learning (DL) shows promise in LDCT reconstruction, it requires large-scale data collection from multiple clients, raising privacy concerns. Federated learning (FL) has been introduced to address these privacy concerns; however, current methods are typically tailored to specific scanning protocols, which limits their generalizability and makes them less effective for unseen protocols. To address these issues, we propose SCAN-PhysFed, a novel SCanning- and ANatomy-level personalized Physics-Driven Federated learning paradigm for LDCT reconstruction. Since the noise distribution in LDCT data is closely tied to scanning protocols and anatomical structures being scanned, we design a dual-level physics-informed way to address these challenges. Specifically, we incorporate physical and anatomical prompts into our physics-informed hypernetworks to capture scanning- and anatomy-specific information, enabling dual-level physics-driven personalization of imaging features. These prompts are derived from the scanning protocol and the radiology report generated by a medical large language model (MLLM), respectively. Subsequently, client-specific decoders project these dual-level personalized imaging features back into the image domain. Besides, to tackle the challenge of unseen data, we introduce a novel protocol vector-quantization strategy (PVQS), which ensures consistent performance across new clients by quantifying the unseen scanning code as one of the codes in the scanning codebook. Extensive experimental results demonstrate the superior performance of SCAN-PhysFed on public datasets.
In egocentric video understanding, the motion of hands and objects as well as their interactions play a significant role by nature. However, existing egocentric video representation learning methods mainly focus on aligning video representation with high-level narrations, overlooking the intricate dynamics between hands and objects. In this work, we aim to integrate the modeling of fine-grained hand-object dynamics into the video representation learning process. Since no suitable data is available, we introduce HOD, a novel pipeline employing a hand-object detector and a large language model to generate high-quality narrations with detailed descriptions of hand-object dynamics. To learn these fine-grained dynamics, we propose EgoVideo, a model with a new lightweight motion adapter to capture fine-grained hand-object motion information. Through our co-training strategy, EgoVideo effectively and efficiently leverages the fine-grained hand-object dynamics in the HOD data. Extensive experiments demonstrate that our method achieves state-of-the-art performance across multiple egocentric downstream tasks, including improvements of 6.3% in EK-100 multi-instance retrieval, 5.7% in EK-100 classification, and 16.3% in EGTEA classification in zero-shot settings. Furthermore, our model exhibits robust generalization capabilities in hand-object interaction and robot manipulation tasks. Code and data are available at https://github.com/OpenRobotLab/EgoHOD/.
Recent open-vocabulary human-object interaction (OV-HOI) detection methods primarily rely on large language model (LLM) for generating auxiliary descriptions and leverage knowledge distilled from CLIP to detect unseen interaction categories. Despite their effectiveness, these methods face two challenges: (1) feature granularity deficiency, due to reliance on last layer visual features for text alignment, leading to the neglect of crucial object-level details from intermediate layers; (2) semantic similarity confusion, resulting from CLIP's inherent biases toward certain classes, while LLM-generated descriptions based solely on labels fail to adequately capture inter-class similarities. To address these challenges, we propose a stratified granular comparison network. First, we introduce a granularity sensing alignment module that aggregates global semantic features with local details, refining interaction representations and ensuring robust alignment between intermediate visual features and text embeddings. Second, we develop a hierarchical group comparison module that recursively compares and groups classes using LLMs, generating fine-grained and discriminative descriptions for each interaction category. Experimental results on two widely-used benchmark datasets, SWIG-HOI and HICO-DET, demonstrate that our method achieves state-of-the-art results in OV-HOI detection. Codes will be released on https://github.com/Phil0212/SGC-Net.
The increasing integration of large language models (LLMs) based conversational agents into everyday life raises critical cognitive and social questions about their potential to influence human opinions. Although previous studies have shown that LLM-based agents can generate persuasive content, these typically involve controlled English-language settings. Addressing this, our preregistered study explored LLMs' persuasive capabilities in more ecological, unconstrained scenarios, examining both static (written paragraphs) and dynamic (conversations via Telegram) interaction types. Conducted entirely in Hebrew with 200 participants, the study assessed the persuasive effects of both LLM and human interlocutors on controversial civil policy topics. Results indicated that participants adopted LLM and human perspectives similarly, with significant opinion changes evident across all conditions, regardless of interlocutor type or interaction mode. Confidence levels increased significantly in most scenarios. These findings demonstrate LLM-based agents' robust persuasive capabilities across diverse sources and settings, highlighting their potential impact on shaping public opinions.
Safety alignment is an important procedure before the official deployment of a Large Language Model (LLM). While safety alignment has been extensively studied for LLM, there is still a large research gap for Large Reasoning Models (LRMs) that equip with improved reasoning capability. We in this paper systematically examine a simplified pipeline for producing safety aligned LRMs. With our evaluation of various LRMs, we deliver two main findings: i) Safety alignment can be done upon the LRM to restore its safety capability. ii) Safety alignment leads to a degradation of the reasoning capability of LRMs. The two findings show that there exists a trade-off between reasoning and safety capability with the sequential LRM production pipeline. The discovered trade-off, which we name Safety Tax, should shed light on future endeavors of safety research on LRMs. As a by-product, we curate a dataset called DirectRefusal, which might serve as an alternative dataset for safety alignment. Our source code is available at https://github.com/git-disl/Safety-Tax.