Visual Question Answering (VQA) requires reasoning across visual and textual modalities, yet Large Vision-Language Models (LVLMs) often lack integrated commonsense knowledge, limiting their robustness in real-world scenarios. To address this, we introduce MAGIC-VQA, a novel framework that enhances VQA by systematically integrating commonsense knowledge with LVLMs. MAGIC-VQA employs a three-stage process: (1) Explicit Knowledge Integration from external sources, (2) By-Type Post-Processing for contextual refinement, and (3) Implicit Knowledge Augmentation using a Graph Neural Network (GNN) for structured reasoning. While GNNs bring greater depth to structured inference, they enable superior relational inference beyond LVLMs. MAGIC-VQA bridges a key gap by unifying commonsensse knowledge with LVLM-driven reasoning, eliminating the need for extensive pre-training or complex prompt tuning. Our framework achieves state-of-the-art performance on benchmark datasets, significantly improving commonsense reasoning in VQA.

Construction Grammar hypothesizes that knowledge of a language consists chiefly of knowledge of form-meaning pairs (''constructions'') that include vocabulary, general grammar rules, and even idiosyncratic patterns. Recent work has shown that transformer language models represent at least some constructional patterns, including ones where the construction is rare overall. In this work, we probe BERT's representation of the form and meaning of a minor construction of English, the NPN (noun-preposition-noun) construction -- exhibited in such expressions as face to face and day to day -- which is known to be polysemous. We construct a benchmark dataset of semantically annotated corpus instances (including distractors that superficially resemble the construction). With this dataset, we train and evaluate probing classifiers. They achieve decent discrimination of the construction from distractors, as well as sense disambiguation among true instances of the construction, revealing that BERT embeddings carry indications of the construction's semantics. Moreover, artificially permuting the word order of true construction instances causes them to be rejected, indicating sensitivity to matters of form. We conclude that BERT does latently encode at least some knowledge of the NPN construction going beyond a surface syntactic pattern and lexical cues.

Large language models (LLMs) have demonstrated impressive capabilities, but their enormous size poses significant challenges for deployment in real-world applications. To address this issue, researchers have sought to apply network pruning techniques to LLMs. A critical challenge in pruning is allocation the sparsity for each layer. Recent sparsity allocation methods is often based on heuristics or search that can easily lead to suboptimal performance. In this paper, we conducted an extensive investigation into various LLMs and revealed three significant discoveries: (1) the layerwise pruning sensitivity (LPS) of LLMs is highly non-uniform, (2) the choice of pruning metric affects LPS, and (3) the performance of a sparse model is related to the uniformity of its layerwise redundancy level. Based on these observations, we propose that the layerwise sparsity of LLMs should adhere to three principles: \emph{non-uniformity}, \emph{pruning metric dependency}, and \emph{uniform layerwise redundancy level} in the pruned model. To this end, we proposed Maximum Redundancy Pruning (MRP), an iterative pruning algorithm that prunes in the most redundant layers (\emph{i.e.}, those with the highest non-outlier ratio) at each iteration. The achieved layerwise sparsity aligns with the outlined principles. We conducted extensive experiments on publicly available LLMs, including the LLaMA2 and OPT, across various benchmarks. Experimental results validate the effectiveness of MRP, demonstrating its superiority over previous methods.

Despite recent leaps forward in image quality, the biases found in videos generated by AI tools, like OpenAI's Sora, are as conspicuous as ever. A WIRED investigation, which included a review of hundreds of AI-generated videos, has found that Sora's model perpetuates sexist, racist, and ableist stereotypes in its results. In Sora's world, everyone is good-looking. Pilots, CEOs, and college professors are men, while flight attendants, receptionists, and childcare workers are women. Disabled people are wheelchair users, interracial relationships are tricky to generate, and fat people don't run.

In this paper, we evaluate the ability of Large Language Models (LLMs) to assess the veracity of claims in ''news reports'' generated by themselves or other LLMs. Our goal is to determine whether LLMs can effectively fact-check their own content, using methods similar to those used to verify claims made by humans. Our findings indicate that LLMs are more effective at assessing claims in national or international news stories than in local news stories, better at evaluating static information than dynamic information, and better at verifying true claims compared to false ones. We hypothesize that this disparity arises because the former types of claims are better represented in the training data. Additionally, we find that incorporating retrieved results from a search engine in a Retrieval-Augmented Generation (RAG) setting significantly reduces the number of claims an LLM cannot assess. However, this approach also increases the occurrence of incorrect assessments, partly due to irrelevant or low-quality search results. This diagnostic study highlights the need for future research on fact-checking machine-generated reports to prioritize improving the precision and relevance of retrieved information to better support fact-checking efforts. Furthermore, claims about dynamic events and local news may require human-in-the-loop fact-checking systems to ensure accuracy and reliability.

Fine-tuning large language models (LLMs) on private, on-device data can empower tailored personalized AI agents. However, fine-tuning LLMs on resource-constrained edge devices faces significant challenges, including excessive computation overhead, device heterogeneity, and data imbalance. This paper proposes SplitFrozen, a split learning framework that enables efficient LLM fine-tuning by strategically freezing device-side model layers while centralizing parameter-efficient fine-tuning on the server. Our framework partitions LLMs into device-side frozen layers and server-side fine-tuning layers, where heterogeneous resource-constrained devices execute only forward propagation. To minimize server-side training costs, we integrate Low-Rank Adaptation (LoRA) into the server-side layers. A pipeline parallelism strategy further optimizes training efficiency by decoupling device-server computations and leveraging decomposed backward propagation. Experiments on GPT-2 with the MRPC, MNLI-matched, and SST-2 datasets demonstrate that SplitFrozen outperforms FedLoRA and SplitLoRA by 69.4\% model accuracy under extremely imbalanced data, while reducing up to 86.8\% device-side computations and 50.2\% total training time. Experiments also validate the scalability of SplitFrozen on content generation task using Llama-3.2 model on GSM8K dataset.

Aligning large language models (LLMs) with human preferences and values is vital for application. However, current alignment methods face three main limitations: (1) reliance on costly human annotation; (2) alignment tax; (3) shallow alignment vulnerable to jailbreak attacks. Additionally, current alignment datasets often suffer from uneven distributions, leading to overrepresentation of some topics and neglect of others. To address these issues, we propose SRMIR (Shadow Reward Models Based on Introspective Reasoning), inspired by shadow models in membership inference attacks. We first construct a balanced safety Chain of Draft (CoD) dataset across 7 harmful types with structured prompt leveraging the introspective reasoning capabilities of LLMs, then train a set of specialized reward models to guide policy optimization through Group Relative Policy Optimization (GRPO). We apply two strategies, linear combination and categorized approach, to integrate shadow reward models for policy optimization. By comparison, we find that the latter achieves superior alignment despite higher computational costs. Experiments across several LLMs demonstrate SRMIR significantly outperforms existing methods. Disclaimer: This paper contains potentially disturbing and offensive content.

T ransforming a dynamic hypothesis into a causal loop diagram (CLD) is crucial for System Dynamics Modelling. Extracting key variables and causal relationships from text to build a CLD is often challenging and time - consuming for novice modelers, limiting SD tool adoption. This paper introduces and tests a method for automating the translation of dynamic hypotheses into CLDs using large language models (LLMs) with curated prompting techniques. We first describe how LLMs work and how they can make the inferences needed to build CLDs using a standard digraph structure. Next, we develop a set of simple dynamic hypothe ses and corresponding CLDs from leading SD textbooks. We then compare the four different combinations of prompting technique s, evaluating their performance against CLD s labeled by expert modelers . Results show that for simple model structures and using curated prompting techniques, LLMs can generate CLDs of a similar quality to expert - built ones, accelerating CLD creation.

Scene graphs capture complex relationships among objects, serving as strong priors for content generation and manipulation. Yet, reasonably manipulating scene graphs -- whether by adding nodes or modifying edges -- remains a challenging and untouched task. Tasks such as adding a node to the graph or reasoning about a node's relationships with all others are computationally intractable, as even a single edge modification can trigger conflicts due to the intricate interdependencies within the graph. To address these challenges, we introduce SG-Tailor, an autoregressive model that predicts the conflict-free relationship between any two nodes. SG-Tailor not only infers inter-object relationships, including generating commonsense edges for newly added nodes but also resolves conflicts arising from edge modifications to produce coherent, manipulated graphs for downstream tasks. For node addition, the model queries the target node and other nodes from the graph to predict the appropriate relationships. For edge modification, SG-Tailor employs a Cut-And-Stitch strategy to solve the conflicts and globally adjust the graph. Extensive experiments demonstrate that SG-Tailor outperforms competing methods by a large margin and can be seamlessly integrated as a plug-in module for scene generation and robotic manipulation tasks.

Abstract--Recent advancements in large language models (LLMs) have catalyzed a substantial surge in demand for LLM services. While traditional cloud-based LLM services satisfy high-accuracy requirements, they fall short in meeting critical demands for low delay and enhanced privacy . T o address these limitations, we propose HA T, a novel device-cloud collaborative inference framework that leverages the complementary strengths of U-shaped inference and speculative decoding. HA T partitions the LLM into three submodels, and the input and output submodels, stacked with a lightweight adapter network, are deployed as a small language model (SLM) on each end device. Meanwhile, the middle submodel, encompassing the majority of the LLM's decoder layers, is hosted in the cloud to perform speculative decoding with on-device SLMs. During inference, HA T exchanges hidden states (rather than raw tokens) of input or draft tokens between devices and the cloud, thereby incurring substantial communication delays. Besides, processing hidden states of long prompts will exacerbate computation delays in the cloud, further compromising inference efficiency . T o improve efficiency, we introduce a prompt chunking mechanism that segments long prompts into shorter chunks, enabling parallel transmission and processing. Furthermore, HA T is implemented to dynamically determine optimal chunk sizes for devices handling long prompts, thereby improving overall inference speed. Extensive experiments are conducted on a physical testbed comprising 30 NVIDIA Jetson devices and a server with 8 NVIDIA A6000 GPUs. Experimental results demonstrate that HA T achieves promising performance improvements, reducing TTFT by 41% to 54% and TBT by 41% to 77% compared to the baselines. Recent advancements in large language models (LLMs) have revolutionized the field of natural language processing, demonstrating unprecedented capabilities across various tasks and triggering exponential growth of LLM services [1], [2]. For instance, OpenAI's ChatGPT provides various services, e.g., chat-based interaction, and automated writing, to approximately 180 million users, and processes over 1.6 billion requests monthly [3]. The underlying architecture of LLM services mainly operates through an autore-gressive process, which involves a prefill phase followed by a decode phase. In prefill phase, the LLM processes all input prompt tokens simultaneously, leveraging parallel computation to generate the initial output token.