arXiv Computer Science @arxiv_cs@qoto.org

Large Language Models (LLMs) exhibit remarkable capabilities but suffer from apparent precision loss, reframed here as information spreading. This reframing shifts the problem from computational precision to an information-theoretic communication issue. We address the K:V and V:K memory problem in LLMs by introducing HDRAM (Holographically Defined Random Access Memory), a symbolic memory framework treating transformer latent space as a spread-spectrum channel. Built upon hypertokens, structured symbolic codes integrating classical error-correcting codes (ECC), holographic computing, and quantum-inspired search, HDRAM recovers distributed information through principled despreading. These phase-coherent memory addresses enable efficient key-value operations and Grover-style search in latent space. By combining ECC grammar with compressed sensing and Krylov subspace alignment, HDRAM significantly improves associative retrieval without architectural changes, demonstrating how Classical-Holographic-Quantum-inspired (CHQ) principles can fortify transformer architectures.

This paper presents NeutroSENSE, a neutrosophic-enhanced ensemble framework for interpretable intrusion detection in IoT environments. By integrating Random Forest, XGBoost, and Logistic Regression with neutrosophic logic, the system decomposes prediction confidence into truth (T), falsity (F), and indeterminacy (I) components, enabling uncertainty quantification and abstention. Predictions with high indeterminacy are flagged for review using both global and adaptive, class-specific thresholds. Evaluated on the IoT-CAD dataset, NeutroSENSE achieved 97% accuracy, while demonstrating that misclassified samples exhibit significantly higher indeterminacy (I = 0.62) than correct ones (I = 0.24). The use of indeterminacy as a proxy for uncertainty enables informed abstention and targeted review-particularly valuable in edge deployments. Figures and tables validate the correlation between I-scores and error likelihood, supporting more trustworthy, human-in-the-loop AI decisions. This work shows that neutrosophic logic enhances both accuracy and explainability, providing a practical foundation for trust-aware AI in edge and fog-based IoT security systems.

Large language models (LLMs) demand considerable computational, energy, and financial resources during both training and deployment. While scaling laws for training have guided much of the field's recent progress, inference costs now represent a significant and growing component of the overall resource burden, particularly for reasoning-focused models. Existing characterizations of compute-optimality that consider model size, dataset size, and inference tokens in isolation or in fixed combinations risk overlooking more efficient operating points. We introduce directed stochastic skill search (DS3), a general framework that represents inference as stochastic traversal over a learned skill graph. From a simplified yet expressive instantiation, we derive closed-form expressions for task success and compute cost across a wide range of inference strategies -- including chain-of-thought (CoT) and tree-of-thought (ToT) -- enabling comparative analysis as a function of task difficulty and model capability. To that end, we extend a prior first-principles tripartite graph framework of LLM training to incorporate inference, and separately bridge DS3 with empirical methods that characterize LLM scaling behavior. We theoretically recover empirically observed patterns, including: linear accuracy scaling with logarithmic compute; variation in preferred inference strategies as a function of task difficulty and model capability; emergent behavior elicited by reasoning even when performance plateaus under parameter scaling; and both best-of-N (BoN) and majority voting behavior captured within a unified analytical framework. By explicitly characterizing training-inference interdependencies, our framework deepens theoretical understanding and supports principled algorithmic design and resource allocation.

Disaster response requires rapid, adaptive decision-making in chaotic environments. SwarmFusion, a novel hybrid framework, integrates particle swarm optimization with convolutional neural networks to optimize real-time resource allocation and path planning. By processing live satellite, drone, and sensor data, SwarmFusion enhances situational awareness and operational efficiency in flood and wildfire scenarios. Simulations using the DisasterSim2025 dataset demonstrate up to 40 percentage faster response times and 90 percentage survivor coverage compared to baseline methods. This scalable, data-driven approach offers a transformative solution for time-critical disaster management, with potential applications across diverse crisis scenarios.

We propose MVGBench, a comprehensive benchmark for multi-view image generation models (MVGs) that evaluates 3D consistency in geometry and texture, image quality, and semantics (using vision language models). Recently, MVGs have been the main driving force in 3D object creation. However, existing metrics compare generated images against ground truth target views, which is not suitable for generative tasks where multiple solutions exist while differing from ground truth. Furthermore, different MVGs are trained on different view angles, synthetic data and specific lightings -- robustness to these factors and generalization to real data are rarely evaluated thoroughly. Without a rigorous evaluation protocol, it is also unclear what design choices contribute to the progress of MVGs. MVGBench evaluates three different aspects: best setup performance, generalization to real data and robustness. Instead of comparing against ground truth, we introduce a novel 3D self-consistency metric which compares 3D reconstructions from disjoint generated multi-views. We systematically compare 12 existing MVGs on 4 different curated real and synthetic datasets. With our analysis, we identify important limitations of existing methods specially in terms of robustness and generalization, and we find the most critical design choices. Using the discovered best practices, we propose ViFiGen, a method that outperforms all evaluated MVGs on 3D consistency. Our code, model, and benchmark suite will be publicly released.

This paper presents a new educational framework for integrating generative artificial intelligence (GenAI) platforms such as ChatGPT, Claude, and Gemini into laboratory activities aimed at developing critical thinking and digital literacy among undergraduate students. Recognizing the limitations and risks of uncritical reliance on large language models (LLMs), the proposed pedagogical model reframes GenAI as a research subject and cognitive tool. Students formulate discipline-specific prompts and evaluate GenAI-generated responses in text, image, and video modalities. A pilot implementation in a general astronomy course for non-science majors demonstrated high levels of engagement and critical reflection, with many students continuing the activity after class and presenting results at a research symposium. The results highlight the importance of structured AI interactions in education and suggest that GenAI can improve learning outcomes when combined with reflective assessment methods. The study proposes a replicable model for interdisciplinary AI-integrated lab work, adaptable to scientific disciplines. See the guide to learning activities based on Generative-Ai platforms: https://doi.org/10.5281/zenodo.15555802

Grounding natural language queries in graphical user interfaces (GUIs) poses unique challenges due to the diversity of visual elements, spatial clutter, and the ambiguity of language. In this paper, we introduce DiMo-GUI, a training-free framework for GUI grounding that leverages two core strategies: dynamic visual grounding and modality-aware optimization. Instead of treating the GUI as a monolithic image, our method splits the input into textual elements and iconic elements, allowing the model to reason over each modality independently using general-purpose vision-language models. When predictions are ambiguous or incorrect, DiMo-GUI dynamically focuses attention by generating candidate focal regions centered on the model's initial predictions and incrementally zooms into subregions to refine the grounding result. This hierarchical refinement process helps disambiguate visually crowded layouts without the need for additional training or annotations. We evaluate our approach on standard GUI grounding benchmarks and demonstrate consistent improvements over baseline inference pipelines, highlighting the effectiveness of combining modality separation with region-focused reasoning.

Efficient elevator traffic management in large buildings is critical for minimizing passenger travel times and energy consumption. Because heuristic- or pattern-detection-based controllers struggle with the stochastic and combinatorial nature of dispatching, we model the six-elevator, fifteen-floor system at Vrije Universiteit Amsterdam as a Markov Decision Process and train an end-to-end Reinforcement Learning (RL) Elevator Group Control System (EGCS). Key innovations include a novel action space encoding to handle the combinatorial complexity of elevator dispatching, the introduction of infra-steps to model continuous passenger arrivals, and a tailored reward signal to improve learning efficiency. In addition, we explore various ways to adapt the discounting factor to the infra-step formulation. We investigate RL architectures based on Dueling Double Deep Q-learning, showing that the proposed RL-based EGCS adapts to fluctuating traffic patterns, learns from a highly stochastic environment, and thereby outperforms a traditional rule-based algorithm.

A deep neural network (DNN) is said to be undistillable if, when used as a black-box input-output teacher, it cannot be distilled through knowledge distillation (KD). In this case, the distilled student (referred to as the knockoff student) does not outperform a student trained independently with label smoothing (LS student) in terms of prediction accuracy. To protect intellectual property of DNNs, it is desirable to build undistillable DNNs. To this end, it is first observed that an undistillable DNN may have the trait that each cluster of its output probability distributions in response to all sample instances with the same label should be highly concentrated to the extent that each cluster corresponding to each label should ideally collapse into one probability distribution. Based on this observation and by measuring the concentration of each cluster in terms of conditional mutual information (CMI), a new training method called CMI minimized (CMIM) method is proposed, which trains a DNN by jointly minimizing the conventional cross entropy (CE) loss and the CMI values of all temperature scaled clusters across the entire temperature spectrum. The resulting CMIM model is shown, by extensive experiments, to be undistillable by all tested KD methods existing in the literature. That is, the knockoff students distilled by these KD methods from the CMIM model underperform the respective LS students. In addition, the CMIM model is also shown to performs better than the model trained with the CE loss alone in terms of their own prediction accuracy.

Forecasting complex time series is an important yet challenging problem that involves various industrial applications. Recently, masked time-series modeling has been proposed to effectively model temporal dependencies for forecasting by reconstructing masked segments from unmasked ones. However, since the semantic information in time series is involved in intricate temporal variations generated by multiple time series components, simply masking a raw time series ignores the inherent semantic structure, which may cause MTM to learn spurious temporal patterns present in the raw data. To capture distinct temporal semantics, we show that masked modeling techniques should address entangled patterns through a decomposition approach. Specifically, we propose ST-MTM, a masked time-series modeling framework with seasonal-trend decomposition, which includes a novel masking method for the seasonal-trend components that incorporates different temporal variations from each component. ST-MTM uses a period masking strategy for seasonal components to produce multiple masked seasonal series based on inherent multi-periodicity and a sub-series masking strategy for trend components to mask temporal regions that share similar variations. The proposed masking method presents an effective pre-training task for learning intricate temporal variations and dependencies. Additionally, ST-MTM introduces a contrastive learning task to support masked modeling by enhancing contextual consistency among multiple masked seasonal representations. Experimental results show that our proposed ST-MTM achieves consistently superior forecasting performance compared to existing masked modeling, contrastive learning, and supervised forecasting methods.

Recent advancements in deep learning have significantly enhanced multilingual automatic speech recognition (ASR) due to the development of advanced model architectures and available large-scale multilingual datasets. Despite that, multilingual ASR still suffers from the curse of multilinguality in that different languages tend to interfere with each other, making it difficult for the ASR model to identify multiple languages effectively while sharing model capacity across them. This paper proposes an efficient finetuning framework for customized multilingual ASR via prepared LoRA language experts based on Whisper. Through LoRA expert fusion or knowledge distillation, our approach achieves better recognition performance on target languages than standard fine-tuning methods. Experimental results demonstrate that the proposed models yield approximately 10\% and 15\% relative performance gains in language-aware and language-agnostic scenarios, respectively.

Multimodal Knowledge Graphs (MMKGs), which represent explicit knowledge across multiple modalities, play a pivotal role by complementing the implicit knowledge of Multimodal Large Language Models (MLLMs) and enabling more grounded reasoning via Retrieval Augmented Generation (RAG). However, existing MMKGs are generally limited in scope: they are often constructed by augmenting pre-existing knowledge graphs, which restricts their knowledge, resulting in outdated or incomplete knowledge coverage, and they often support only a narrow range of modalities, such as text and visual information. These limitations reduce their extensibility and applicability to a broad range of multimodal tasks, particularly as the field shifts toward richer modalities such as video and audio in recent MLLMs. Therefore, we propose the Visual-Audio-Text Knowledge Graph (VAT-KG), the first concept-centric and knowledge-intensive multimodal knowledge graph that covers visual, audio, and text information, where each triplet is linked to multimodal data and enriched with detailed descriptions of concepts. Specifically, our construction pipeline ensures cross-modal knowledge alignment between multimodal data and fine-grained semantics through a series of stringent filtering and alignment steps, enabling the automatic generation of MMKGs from any multimodal dataset. We further introduce a novel multimodal RAG framework that retrieves detailed concept-level knowledge in response to queries from arbitrary modalities. Experiments on question answering tasks across various modalities demonstrate the effectiveness of VAT-KG in supporting MLLMs, highlighting its practical value in unifying and leveraging multimodal knowledge.

The rapid spread of fake news across multimedia platforms presents serious challenges to information credibility. In this paper, we propose a Debunk-and-Infer framework for Fake News Detection(DIFND) that leverages debunking knowledge to enhance both the performance and interpretability of fake news detection. DIFND integrates the generative strength of conditional diffusion models with the collaborative reasoning capabilities of multimodal large language models (MLLMs). Specifically, debunk diffusion is employed to generate refuting or authenticating evidence based on the multimodal content of news videos, enriching the evaluation process with diverse yet semantically aligned synthetic samples. To improve inference, we propose a chain-of-debunk strategy where a multi-agent MLLM system produces logic-grounded, multimodal-aware reasoning content and final veracity judgment. By jointly modeling multimodal features, generative debunking cues, and reasoning-rich verification within a unified architecture, DIFND achieves notable improvements in detection accuracy. Extensive experiments on the FakeSV and FVC datasets show that DIFND not only outperforms existing approaches but also delivers trustworthy decisions.

Forecasting is a challenging task that offers a clearly measurable way to study AI systems. Forecasting requires a large amount of research on the internet, and evaluations require time for events to happen, making the development of forecasting benchmarks challenging. To date, no forecasting benchmark provides a realistic, hermetic, and repeatable environment for LLM forecasters. We introduce Bench To the Future (BTF), a "pastcasting" benchmark with hundreds of high-quality questions for which the resolution is already known. Each question is accompanied by a large offline corpus of tens of thousands of relevant web pages, enabling a way to elicit realistic "forecasts" on past events from LLMs. Results suggest that our pastcasting environment can produce results comparable to those based on forecasts using the internet on at-the-time unresolved questions. We show results benchmarking agent and chain-of-thought forecasting approaches using several LLMs, including the recently-released Claude 4 models, and demonstrate BTF's ability to track steady forecasting capability progress over time. We intend this to be a living benchmark, with new questions added continually to account for increasing training data cutoff dates. We invite researchers to contact us at hello@futuresearch.ai to utilize our benchmark or tooling for their own research.

Large language models (LLMs) have demonstrated their strong capabilities in various domains, and have been recently integrated for graph analysis as graph language models (GLMs). With LLMs as the predictor, some GLMs can interpret unseen tasks described by natural language, and learn from a few examples in the prompts without parameter tuning, known as in-context learning (ICL). Another subset of GLMs utilizes abundant training labels to enhance model performance, known as instruction tuning. However, we argue that ICL on graphs has effectiveness issues due to fixed parameters and efficiency issues due to long context. Meanwhile, the large amount of labeled data required for instruction tuning can be difficult to obtain in real-world scenarios. To this end, we aim to introduce an extra parameter adaptation stage that can efficiently tailor GLMs to an unseen graph and task with only a few labeled examples, in exchange for better prediction accuracy and faster inference speed. For implementation, in this paper we propose GraphLAMA method, with its model backbone and learning schemes specialized for efficient tuning and inference. Specifically, for model backbone, we use a graph neural network (GNN) with several well-designed components to transform nodes into the representation space of LLM tokens. Task instructions can then be represented as a mixture of node and language tokens. In the pre-training stage, model parameters except the LLM will be trained with different tasks to capture general knowledge. In the adaptation stage, only a few pre-trained parameters will be updated based on few-shot examples. Extensive experiments on few/zero-shot node classification and summary generation show that our proposed GraphLAMA achieves state-of-the-art performance with 4.91% absolution improvement in accuracy. Compared with ICL, our inference speed can be 10 times faster under 5-shot setting.

This study investigates the effectiveness of reinforcement learning (RL) fine-tuning techniques on a compact language model (Qwen2.5-0.5B Base) for two challenging tasks: instruction following and mathematical reasoning. We compare supervised fine-tuning (SFT), Direct Preference Optimization (DPO) using preference-labeled data, and Reinforce Leave-One-Out (RLOO) with reward models. Our experiments show that RLOO with DeBERTa reward modeling achieves the best alignment, while DPO provides strong and consistent results. For math reasoing tasks, synthetic data augmentation and best-of-N sampling with an external verifier significantly improve accuracy, showing the potential of combining fine-tuning with inference-time tools. This study highlights key trade-offs and practical strategies for training lightweight, task-aligned small-scale language models.

Despite their widespread use in fact-checking, moderation, and high-stakes decision-making, large language models (LLMs) remain poorly understood as judges of truth. This study presents the largest evaluation to date of LLMs' veracity detection capabilities and the first analysis of these capabilities in reasoning models. We had eight LLMs make 4,800 veracity judgments across several prompts, comparing reasoning and non-reasoning models. We find that rates of truth-bias, or the likelihood to believe a statement is true, regardless of whether it is actually true, are lower in reasoning models than in non-reasoning models, but still higher than human benchmarks. Most concerning, we identify sycophantic tendencies in several advanced models (o4-mini and GPT-4.1 from OpenAI, R1 from DeepSeek), which displayed an asymmetry in detection accuracy, performing well in truth accuracy but poorly in deception accuracy. This suggests that capability advances alone do not resolve fundamental veracity detection challenges in LLMs.

In the architectural design process, floor plan generation is inherently progressive and iterative. However, existing generative models for floor plans are predominantly end-to-end generation that produce an entire pixel-based layout in a single pass. This paradigm is often incompatible with the incremental workflows observed in real-world architectural practice. To address this issue, we draw inspiration from the autoregressive 'next token prediction' mechanism commonly used in large language models, and propose a novel 'next room prediction' paradigm tailored to architectural floor plan modeling. Experimental evaluation indicates that FPDS demonstrates competitive performance in comparison to diffusion models and Tell2Design in the text-to-floorplan task, indicating its potential applicability in supporting future intelligent architectural design.

While large language models (LLMs) have demonstrated impressive performance across a wide range of natural language processing (NLP) tasks in high-resource languages, their capabilities in low-resource and minority languages remain significantly underexplored. Formosan languages -- a subgroup of Austronesian languages spoken in Taiwan -- are both linguistically rich and endangered, largely due to the sociolinguistic dominance of Mandarin. In this work, we introduce FORMOSANBENCH, the first benchmark for evaluating LLMs on low-resource Austronesian languages. It covers three endangered Formosan languages: Atayal, Amis, and Paiwan, across three core NLP tasks: machine translation, automatic speech recognition (ASR), and text summarization. We assess model performance in zero-shot, 10-shot, and fine-tuned settings using FORMOSANBENCH. Our results reveal a substantial performance gap between high-resource and Formosan languages. Existing LLMs consistently underperform across all tasks, with 10-shot learning and fine-tuning offering only limited improvements. These findings underscore the urgent need for more inclusive NLP technologies that can effectively support endangered and underrepresented languages. We release our datasets and code to facilitate future research in this direction.