Large Multimodal Models (LMMs) demonstrate significant cross-modal reasoning capabilities. However, financial applications face challenges due to the lack of high-quality multimodal reasoning datasets and the inefficiency of existing training paradigms for reasoning enhancement. To address these issues, we propose an integrated framework, FinLMM-R1, combining an automated and scalable pipeline for data construction with enhanced training strategies to improve the multimodal reasoning of LMM. The Automated and Scalable Pipeline (ASP) resolves textual-visual misalignment in financial reports through a separate paradigm of question-answer generation and image-question alignment, ensuring data integrity and extraction efficiency. Through ASP, we collect 89,378 aligned image-question pairs from 23,397 financial reports, covering tasks such as arithmetic reasoning, statistics reasoning, financial explanation, and financial knowledge. Moreover, we introduce the Thinking with Adversarial Reward in LMM (TAR-LMM), extending the prior two-stage training framework [1] with additional reward mechanisms. In the first stage, we focus on text-only tasks with format and accuracy rewards to guide the model in generating well-structured thinking contents. In the second stage, we construct multi-image contrastive samples with additional reward components including image selection, thinking content length, and adversarial reward to jointly optimize the LMM across visual perception, reasoning efficiency, and logical coherence. Extensive experiments on 7 benchmarks show ASP-derived dataset and training framework significantly improve answer accuracy and reasoning depth over existing reasoning LMMs in both general and financial multimodal contexts.
Language models can retain dangerous knowledge and skills even after extensive safety fine-tuning, posing both misuse and misalignment risks. Recent studies show that even specialized unlearning methods can be easily reversed. To address this, we systematically evaluate many existing and novel components of unlearning methods and identify ones crucial for irreversible unlearning. We introduce Disruption Masking, a technique in which we only allow updating weights, where the signs of the unlearning gradient and the retaining gradient are the same. This ensures all updates are non-disruptive. Additionally, we identify the need for normalizing the unlearning gradients, and also confirm the usefulness of meta-learning. We combine these insights into MUDMAN (Meta-Unlearning with Disruption Masking and Normalization) and validate its effectiveness at preventing the recovery of dangerous capabilities. MUDMAN outperforms the prior TAR method by 40\%, setting a new state-of-the-art for robust unlearning.
Modern multimodal large language models (MLLMs) can reason over hour-long video, yet their key-value (KV) cache grows linearly with time--quickly exceeding the fixed memory of phones, AR glasses, and edge robots. Prior compression schemes either assume the whole video and user query are available offline or must first build the full cache, so memory still scales with stream length. InfiniPot-V is the first training-free, query-agnostic framework that enforces a hard, length-independent memory cap for streaming video understanding. During video encoding it monitors the cache and, once a user-set threshold is reached, runs a lightweight compression pass that (i) removes temporally redundant tokens via Temporal-axis Redundancy (TaR) metric and (ii) keeps semantically significant tokens via Value-Norm (VaN) ranking. Across four open-source MLLMs and four long-video and two streaming-video benchmarks, InfiniPot-V cuts peak GPU memory by up to 94%, sustains real-time generation, and matches or surpasses full-cache accuracy--even in multi-turn dialogues. By dissolving the KV cache bottleneck without retraining or query knowledge, InfiniPot-V closes the gap for on-device streaming video assistants.
This paper presents a multimodal framework that attempts to unify visual understanding and generation within a shared discrete semantic representation. At its core is the Text-Aligned Tokenizer (TA-Tok), which converts images into discrete tokens using a text-aligned codebook projected from a large language model's (LLM) vocabulary. By integrating vision and text into a unified space with an expanded vocabulary, our multimodal LLM, Tar, enables cross-modal input and output through a shared interface, without the need for modality-specific designs. Additionally, we propose scale-adaptive encoding and decoding to balance efficiency and visual detail, along with a generative de-tokenizer to produce high-fidelity visual outputs. To address diverse decoding needs, we utilize two complementary de-tokenizers: a fast autoregressive model and a diffusion-based model. To enhance modality fusion, we investigate advanced pre-training tasks, demonstrating improvements in both visual understanding and generation. Experiments across benchmarks show that Tar matches or surpasses existing multimodal LLM methods, achieving faster convergence and greater training efficiency. Code, models, and data are available at https://tar.csuhan.com
In this work, we show that some machine unlearning methods may fail when subjected to straightforward prompt attacks. We systematically evaluate eight unlearning techniques across three model families, and employ output-based, logit-based, and probe analysis to determine to what extent supposedly unlearned knowledge can be retrieved. While methods like RMU and TAR demonstrate robust unlearning, ELM remains vulnerable to specific prompt attacks (e.g., Hindi filler text in original prompt recovering 57.3% accuracy). Our logit analysis also confirms that unlearned models are generally not hiding knowledge by modifying the way the answer is formatted, as the correlation between output and logit accuracy is strong. These results challenge prevailing assumptions about unlearning effectiveness and highlight the need for evaluation frameworks that can reliably distinguish between true knowledge removal and superficial output suppression. We also publicly make available our evaluation framework to easily evaluate prompting techniques to retrieve unlearning knowledge.
Actor-Critic methods are widely used for their scalability, yet existing theoretical guarantees for infinite-horizon average-reward Markov Decision Processes (MDPs) often rely on restrictive ergodicity assumptions. We propose NAC-B, a Natural Actor-Critic with Batching, that achieves order-optimal regret of $\tilde{O}(\sqrt{T})$ in infinite-horizon average-reward MDPs under the unichain assumption, which permits both transient states and periodicity. This assumption is among the weakest under which the classic policy gradient theorem remains valid for average-reward settings. NAC-B employs function approximation for both the actor and the critic, enabling scalability to problems with large state and action spaces. The use of batching in our algorithm helps mitigate potential periodicity in the MDP and reduces stochasticity in gradient estimates, and our analysis formalizes these benefits through the introduction of the constants $C_{\text{hit}}$ and $C_{\text{tar}}$, which characterize the rate at which empirical averages over Markovian samples converge to the stationary distribution.
Multi-agent collaboration has emerged as a pivotal paradigm for addressing complex, distributed tasks in large language model (LLM)-driven applications. While prior research has focused on high-level architectural frameworks, the granular mechanisms governing agents, critical to performance and scalability, remain underexplored. This study systematically investigates four dimensions of collaboration strategies: (1) agent governance, (2) participation control, (3) interaction dynamics, and (4) dialogue history management. Through rigorous experimentation under two context-dependent scenarios: Distributed Evidence Integration (DEI) and Structured Evidence Synthesis (SES), we quantify the impact of these strategies on both task accuracy and computational efficiency. Our findings reveal that centralized governance, instructor-led participation, ordered interaction patterns, and instructor-curated context summarization collectively optimize the trade-off between decision quality and resource utilization with the support of the proposed Token-Accuracy Ratio (TAR). This work establishes a foundation for designing adaptive, scalable multi-agent systems, shifting the focus from structural novelty to strategic interaction mechanics.
Despite recent advances in facial recognition, there remains a fundamental issue concerning degradations in performance due to substantial perspective (pose) differences between enrollment and query (probe) imagery. Therefore, we propose a novel domain adaptive framework to facilitate improved performances across large discrepancies in pose by enabling image-based (2D) representations to infer properties of inherently pose invariant point cloud (3D) representations. Specifically, our proposed framework achieves better pose invariance by using (1) a shared (joint) attention mapping to emphasize common patterns that are most correlated between 2D facial images and 3D facial data and (2) a joint entropy regularizing loss to promote better consistency$\unicode{x2014}$enhancing correlations among the intersecting 2D and 3D representations$\unicode{x2014}$by leveraging both attention maps. This framework is evaluated on FaceScape and ARL-VTF datasets, where it outperforms competitive methods by achieving profile (90$\unicode{x00b0}$$\unicode{x002b}$) TAR @ 1$\unicode{x0025}$ FAR improvements of at least 7.1$\unicode{x0025}$ and 1.57$\unicode{x0025}$, respectively.
We address the problem of whole-body person recognition in unconstrained environments. This problem arises in surveillance scenarios such as those in the IARPA Biometric Recognition and Identification at Altitude and Range (BRIAR) program, where biometric data is captured at long standoff distances, elevated viewing angles, and under adverse atmospheric conditions (e.g., turbulence and high wind velocity). To this end, we propose FarSight, a unified end-to-end system for person recognition that integrates complementary biometric cues across face, gait, and body shape modalities. FarSight incorporates novel algorithms across four core modules: multi-subject detection and tracking, recognition-aware video restoration, modality-specific biometric feature encoding, and quality-guided multi-modal fusion. These components are designed to work cohesively under degraded image conditions, large pose and scale variations, and cross-domain gaps. Extensive experiments on the BRIAR dataset, one of the most comprehensive benchmarks for long-range, multi-modal biometric recognition, demonstrate the effectiveness of FarSight. Compared to our preliminary system, this system achieves a 34.1% absolute gain in 1:1 verification accuracy ([email omitted]% FAR), a 17.8% increase in closed-set identification (Rank-20), and a 34.3% reduction in open-set identification errors (FNIR@1% FPIR). Furthermore, FarSight was evaluated in the 2025 NIST RTE Face in Video Evaluation (FIVE), which conducts standardized face recognition testing on the BRIAR dataset. These results establish FarSight as a state-of-the-art solution for operational biometric recognition in challenging real-world conditions.
A continuing frustration for origin of life scientists is that abiotic and, by extension, pre-biotic attempts to develop self-sustaining, evolving molecular systems tend to produce more dead-end substances than macromolecular products with the necessary potential for biostructure and function -- the so-called `tar problem'. Nevertheless primordial life somehow emerged despite that presumed handicap. A~resolution of this problem is important in emergence-of-life science because it would provide valuable guidance in choosing subsequent paths of investigation, such as identifying pre-biotic patterns on Mars. To study the problem we set up a simple non-equilibrium flow dynamical model for the coupled temperature and mass dynamics of the decomposition of a polymeric carbohydrate adsorbed on a mineral surface, with incident stochastic thermal fluctuations. Results show that the model system behaves as a reciprocating thermochemical oscillator. The output fluctuation distribution is bimodal, with a right-weighted component that guarantees a bias towards detachment and desorption of monomeric species such as ribose, even while tar is formed concomitantly. This fluctuating thermochemical reciprocator may ensure that non-performing polymers can be fractionated into a refractory carbon reservoir and active monomers which may be reincorporated into better-performing polymers with less vulnerability towards adsorptive tarring.
DNA cloning methods are fundamental tools in molecular biology, synthetic biology, and genetic engineering that enable precise DNA manipulation for various scientific and biotechnological applications. This review systematically summarizes the major restriction-free overlapping sequence cloning (RFOSC) techniques currently used in synthetic biology and examines their development, efficiency, practicality, and specific applications. In vitro methods, including Gibson Assembly, Circular Polymerase Extension Cloning (CPEC), Polymerase Incomplete Primer Extension (PIPE), Overlap Extension Cloning (OEC), Flap Endonuclease Cloning (FEN-Cloning), and commercially available techniques such as TOPO and In-Fusion, have been discussed alongside hybrid approaches such as Ligation-Independent Cloning (LIC), Sequence-Independent Cloning (SLIC), and T5 Exonuclease-Dependent Assembly (TEDA). Additionally, in vivo methods leveraging host recombination machinery, including Yeast Homologous Recombination (YHR), In Vivo Assembly (IVA), Transformation-Associated Recombination (TAR), and innovative approaches such as Multiple-Round In Vivo Site-Specific Assembly (MISSA) and Phage Enzyme-Assisted Direct Assembly (PEDA), are critically evaluated. The review highlights that method selection should consider the scale, complexity, cost, and specific needs of individual research projects, noting that no single technique is universally optimal. Future trends suggest the increased integration of enzymatic efficiency, host versatility, and automation, broadening the accessibility and capabilities of DNA assembly technologies.
This paper presents a Technology Assisted Review (TAR) stopping approach based on Reinforcement Learning (RL). Previous such approaches offered limited control over stopping behaviour, such as fixing the target recall and tradeoff between preferring to maximise recall or cost. These limitations are overcome by introducing a novel RL environment, GRLStop, that allows a single model to be applied to multiple target recalls, balances the recall/cost tradeoff and integrates a classifier. Experiments were carried out on six benchmark datasets (CLEF e-Health datasets 2017-9, TREC Total Recall, TREC Legal and Reuters RCV1) at multiple target recall levels. Results showed that the proposed approach to be effective compared to multiple baselines in addition to offering greater flexibility.
Aiming to reduce the computational cost of Softmax in massive label space of Face Recognition (FR) benchmarks, recent studies estimate the output using a subset of identities. Although promising, the association between the computation cost and the number of identities in the dataset remains linear only with a reduced ratio. A shared characteristic among available FR methods is the employment of atomic scalar labels during training. Consequently, the input to label matching is through a dot product between the feature vector of the input and the Softmax centroids. Inspired by generative modeling, we present a simple yet effective method that substitutes scalar labels with structured identity code, i.e., a sequence of integers. Specifically, we propose a tokenization scheme that transforms atomic scalar labels into structured identity codes. Then, we train an FR backbone to predict the code for each input instead of its scalar label. As a result, the associated computational cost becomes logarithmic w.r.t. number of identities. We demonstrate the benefits of the proposed method by conducting experiments. In particular, our method outperforms its competitors by 1.52%, and 0.6% at TAR@FAR$=1e-4$ on IJB-B and IJB-C, respectively, while transforming the association between computational cost and the number of identities from linear to logarithmic. See code at https://github.com/msed-Ebrahimi/GIF
Medical Visual Question Answering (MVQA) systems can interpret medical images in response to natural language queries. However, linguistic variability in question phrasing often undermines the consistency of these systems. To address this challenge, we propose a Semantically Equivalent Question Augmentation (SEQA) framework, which leverages large language models (LLMs) to generate diverse yet semantically equivalent rephrasings of questions. Specifically, this approach enriches linguistic diversity while preserving semantic meaning. We further introduce an evaluation metric, Total Agreement Rate with Semantically Equivalent Input and Correct Answer (TAR-SC), which assesses a model's capability to generate consistent and correct responses to semantically equivalent linguistic variations. In addition, we also propose three other diversity metrics - average number of QA items per image (ANQI), average number of questions per image with the same answer (ANQA), and average number of open-ended questions per image with the same semantics (ANQS). Using the SEQA framework, we augmented the benchmarked MVQA public datasets of SLAKE, VQA-RAD, and PathVQA. As a result, all three datasets achieved significant improvements by incorporating more semantically equivalent questions: ANQI increased by an average of 86.1, ANQA by 85.1, and ANQS by 46. Subsequent experiments evaluate three MVQA models (M2I2, MUMC, and BiomedGPT) under both zero-shot and fine-tuning settings on the enhanced datasets. Experimental results in MVQA datasets show that fine-tuned models achieve an average accuracy improvement of 19.35%, while our proposed TAR-SC metric shows an average improvement of 11. 61%, indicating a substantial enhancement in model consistency.
Quadrupedal locomotion via Reinforcement Learning (RL) is commonly addressed using the teacher-student paradigm, where a privileged teacher guides a proprioceptive student policy. However, key challenges such as representation misalignment between the privileged teacher and the proprioceptive-only student, covariate shift due to behavioral cloning, and lack of deployable adaptation lead to poor generalization in real-world scenarios. We propose Teacher-Aligned Representations via Contrastive Learning (TAR), a framework that leverages privileged information with self-supervised contrastive learning to bridge this gap. By aligning representations to a privileged teacher in simulation via contrastive objectives, our student policy learns structured latent spaces and exhibits robust generalization to Out-of-Distribution (OOD) scenarios, surpassing the fully privileged "Teacher". Results showed accelerated training by 2x compared to state-of-the-art baselines to achieve peak performance. OOD scenarios showed better generalization by 40 percent on average compared to existing methods. Additionally, TAR transitions seamlessly into learning during deployment without requiring privileged states, setting a new benchmark in sample-efficient, adaptive locomotion and enabling continual fine-tuning in real-world scenarios. Open-source code and videos are available at https://ammousa.github.io/TARLoco/.
Generative models in Autonomous Driving (AD) enable diverse scene creation, yet existing methods fall short by only capturing a limited range of modalities, restricting the capability of generating controllable scenes for comprehensive evaluation of AD systems. In this paper, we introduce a multimodal generation framework that incorporates four major data modalities, including a novel addition of map modality. With tokenized modalities, our scene sequence generation framework autoregressively predicts each scene while managing computational demands through a two-stage approach. The Temporal AutoRegressive (TAR) component captures inter-frame dynamics for each modality while the Ordered AutoRegressive (OAR) component aligns modalities within each scene by sequentially predicting tokens in a fixed order. To maintain coherence between map and ego-action modalities, we introduce the Action-aware Map Alignment (AMA) module, which applies a transformation based on the ego-action to maintain coherence between these modalities. Our framework effectively generates complex, realistic driving scenes over extended sequences, ensuring multimodal consistency and offering fine-grained control over scene elements. Project page: https://yanhaowu.github.io/UMGen/
In cooperative multi-agent reinforcement learning (MARL), learning effective policies is challenging when global rewards are sparse and delayed. This difficulty arises from the need to assign credit across both agents and time steps, a problem that existing methods often fail to address in episodic, long-horizon tasks. We propose Temporal-Agent Reward Redistribution $TAR^2$, a novel approach that decomposes sparse global rewards into agent-specific, time-step-specific components, thereby providing more frequent and accurate feedback for policy learning. Theoretically, we show that $TAR^2$ (i) aligns with potential-based reward shaping, preserving the same optimal policies as the original environment, and (ii) maintains policy gradient update directions identical to those under the original sparse reward, ensuring unbiased credit signals. Empirical results on two challenging benchmarks, SMACLite and Google Research Football, demonstrate that $TAR^2$ significantly stabilizes and accelerates convergence, outperforming strong baselines like AREL and STAS in both learning speed and final performance. These findings establish $TAR^2$ as a principled and practical solution for agent-temporal credit assignment in sparse-reward multi-agent systems.
In multi-agent environments, agents often struggle to learn optimal policies due to sparse or delayed global rewards, particularly in long-horizon tasks where it is challenging to evaluate actions at intermediate time steps. We introduce Temporal-Agent Reward Redistribution (TAR$^2$), a novel approach designed to address the agent-temporal credit assignment problem by redistributing sparse rewards both temporally and across agents. TAR$^2$ decomposes sparse global rewards into time-step-specific rewards and calculates agent-specific contributions to these rewards. We theoretically prove that TAR$^2$ is equivalent to potential-based reward shaping, ensuring that the optimal policy remains unchanged. Empirical results demonstrate that TAR$^2$ stabilizes and accelerates the learning process. Additionally, we show that when TAR$^2$ is integrated with single-agent reinforcement learning algorithms, it performs as well as or better than traditional multi-agent reinforcement learning methods.
Iris recognition is widely acknowledged for its exceptional accuracy in biometric authentication, traditionally relying on near-infrared (NIR) imaging. Recently, visible spectrum (VIS) imaging via accessible smartphone cameras has been explored for biometric capture. However, a thorough study of iris recognition using smartphone-captured 'High-Quality' VIS images and cross-spectral matching with previously enrolled NIR images has not been conducted. The primary challenge lies in capturing high-quality biometrics, a known limitation of smartphone cameras. This study introduces a novel Android application designed to consistently capture high-quality VIS iris images through automated focus and zoom adjustments. The application integrates a YOLOv3-tiny model for precise eye and iris detection and a lightweight Ghost-Attention U-Net (G-ATTU-Net) for segmentation, while adhering to ISO/IEC 29794-6 standards for image quality. The approach was validated using smartphone-captured VIS and NIR iris images from 47 subjects, achieving a True Acceptance Rate (TAR) of 96.57% for VIS images and 97.95% for NIR images, with consistent performance across various capture distances and iris colors. This robust solution is expected to significantly advance the field of iris biometrics, with important implications for enhancing smartphone security.
The newly introduced Visual State Space Model (VMamba), which employs \textit{State Space Mechanisms} (SSM) to interpret images as sequences of patches, has shown exceptional performance compared to Vision Transformers (ViT) across various computer vision tasks. However, recent studies have highlighted that deep models are susceptible to adversarial attacks. One common approach is to embed a trigger in the training data to retrain the model, causing it to misclassify data samples into a target class, a phenomenon known as a backdoor attack. In this paper, we first evaluate the robustness of the VMamba model against existing backdoor attacks. Based on this evaluation, we introduce a novel architectural backdoor attack, termed BadScan, designed to deceive the VMamba model. This attack utilizes bit plane slicing to create visually imperceptible backdoored images. During testing, if a trigger is detected by performing XOR operations between the $k^{th}$ bit planes of the modified triggered patches, the traditional 2D selective scan (SS2D) mechanism in the visual state space (VSS) block of VMamba is replaced with our newly designed BadScan block, which incorporates four newly developed scanning patterns. We demonstrate that the BadScan backdoor attack represents a significant threat to visual state space models and remains effective even after complete retraining from scratch. Experimental results on two widely used image classification datasets, CIFAR-10, and ImageNet-1K, reveal that while visual state space models generally exhibit robustness against current backdoor attacks, the BadScan attack is particularly effective, achieving a higher Triggered Accuracy Ratio (TAR) in misleading the VMamba model and its variants.