Knowledge-Based Visual Question Answering (KB-VQA) methods focus on tasks that demand reasoning with information extending beyond the explicit content depicted in the image. Early methods relied on explicit knowledge bases to provide this auxiliary information. Recent approaches leverage Large Language Models (LLMs) as implicit knowledge sources. While KB-VQA methods have demonstrated promising results, their potential remains constrained as the auxiliary text provided may not be relevant to the question context, and may also include irrelevant information that could misguide the answer predictor. We introduce a novel four-stage framework called Grounding Caption-Guided Knowledge-Based Visual Question Answering (GC-KBVQA), which enables LLMs to effectively perform zero-shot VQA tasks without the need for end-to-end multimodal training. Innovations include grounding question-aware caption generation to move beyond generic descriptions and have compact, yet detailed and context-rich information. This is combined with knowledge from external sources to create highly informative prompts for the LLM. GC-KBVQA can address a variety of VQA tasks, and does not require task-specific fine-tuning, thus reducing both costs and deployment complexity by leveraging general-purpose, pre-trained LLMs. Comparison with competing KB-VQA methods shows significantly improved performance. Our code will be made public.

With the rise of generative AI (GenAI), Large Language Models are increasingly employed for code generation, becoming active co-authors alongside human programmers. Focusing specifically on this application domain, this paper articulates distinct ``Architectures of Error'' to ground an epistemic distinction between human and machine code generation. Examined through their shared vulnerability to error, this distinction reveals fundamentally different causal origins: human-cognitive versus artificial-stochastic. To develop this framework and substantiate the distinction, the analysis draws critically upon Dennett's mechanistic functionalism and Rescher's methodological pragmatism. I argue that a systematic differentiation of these error profiles raises critical philosophical questions concerning semantic coherence, security robustness, epistemic limits, and control mechanisms in human-AI collaborative software development. The paper also utilizes Floridi's levels of abstraction to provide a nuanced understanding of how these error dimensions interact and may evolve with technological advancements. This analysis aims to offer philosophers a structured framework for understanding GenAI's unique epistemological challenges, shaped by these architectural foundations, while also providing software engineers a basis for more critically informed engagement.

Transformer models power many AI applications but suffer from high inference latency, limiting their use in real-time settings. Multi-device inference can reduce latency by parallelizing computation. Yet, existing methods require high inter-device bandwidth, making them impractical for bandwidth-constrained environments. We propose ASTRA, a communication-efficient framework that accelerates Transformer inference through a novel integration of sequence parallelism and a Mixed-Precision Attention mechanism designed to minimize inter-device communication. ASTRA compresses non-local token embeddings via vector quantization and preserves task accuracy through two optimizations, Noise-Augmented Quantization and Distributed Class Tokens. Experiments on ViT and GPT2 across vision and NLP tasks show that ASTRA achieves up to 2.64X speedups over single-device inference and up to 15.25X speedups over state-of-the-art multi-device inferences, while operating under bandwidths as low as 10 Mbps. ASTRA is open-sourced at https://github.com/xl1990/Astra.

Large language models (LLMs) obtain state of the art zero shot relevance ranking performance on a variety of information retrieval tasks. The two most common prompts to elicit LLM relevance judgments are pointwise scoring (a.k.a. relevance generation), where the LLM sees a single query-document pair and outputs a single relevance score, and listwise ranking (a.k.a. permutation generation), where the LLM sees a query and a list of documents and outputs a permutation, sorting the documents in decreasing order of relevance. The current research community consensus is that listwise ranking yields superior performance, and significant research effort has been devoted to crafting LLM listwise ranking algorithms. The underlying hypothesis is that LLMs are better at making relative relevance judgments than absolute ones. In tension with this hypothesis, we find that the gap between pointwise scoring and listwise ranking shrinks when pointwise scoring is implemented using a sufficiently large ordinal relevance label space, becoming statistically insignificant for many LLM-benchmark dataset combinations (where ``significant'' means ``95\% confidence that listwise ranking improves NDCG@10''). Our evaluations span four LLMs, eight benchmark datasets from the BEIR and TREC-DL suites, and two proprietary datasets with relevance labels collected after the training cut-off of all LLMs evaluated.

Current evaluations of Large Language Model (LLM) steering techniques focus on task-specific performance, overlooking how well steered representations align with human cognition. Using a well-established triadic similarity judgment task, we assessed steered LLMs on their ability to flexibly judge similarity between concepts based on size or kind. We found that prompt-based steering methods outperformed other methods both in terms of steering accuracy and model-to-human alignment. We also found LLMs were biased towards 'kind' similarity and struggled with 'size' alignment. This evaluation approach, grounded in human cognition, adds further support to the efficacy of prompt-based steering and reveals privileged representational axes in LLMs prior to steering.

Current debiasing approaches often result a degradation in model capabilities such as factual accuracy and knowledge retention. Through systematic evaluation across multiple benchmarks, we demonstrate that existing debiasing methods face fundamental trade-offs, particularly in smaller models, leading to reduced truthfulness, knowledge loss, or unintelligible outputs. To address these limitations, we propose a contrastive learning framework that learns through carefully constructed positive and negative examples. Our approach introduces contrast computation and dynamic loss scaling to balance bias mitigation with faithfulness preservation. Experimental results across multiple model scales demonstrate that our method achieves substantial improvements in both toxicity reduction and faithfulness preservation. Most importantly, we show that our framework is the first to consistently improve both metrics simultaneously, avoiding the capability degradation characteristic of existing approaches. These results suggest that explicit modeling of both positive and negative examples through contrastive learning could be a promising direction for reducing the alignment tax in language model debiasing.

Despite significant advances in quality and complexity of the generations in text-to-image models, prompting does not always lead to the desired outputs. Controlling model behaviour by directly steering intermediate model activations has emerged as a viable alternative allowing to reach concepts in latent space that may otherwise remain inaccessible by prompt. In this work, we introduce a set of experiments to deepen our understanding of concept reachability. We design a training data setup with three key obstacles: scarcity of concepts, underspecification of concepts in the captions, and data biases with tied concepts. Our results show: (i) concept reachability in latent space exhibits a distinct phase transition, with only a small number of samples being sufficient to enable reachability, (ii) where in the latent space the intervention is performed critically impacts reachability, showing that certain concepts are reachable only at certain stages of transformation, and (iii) while prompting ability rapidly diminishes with a decrease in quality of the dataset, concepts often remain reliably reachable through steering. Model providers can leverage this to bypass costly retraining and dataset curation and instead innovate with user-facing control mechanisms.

Integrating multiple (sub-)systems is essential to create advanced Information Systems. Difficulties mainly arise when integrating dynamic environments, e.g., the integration at design time of not yet existing services. This has been traditionally addressed using a registry that provides the API documentation of the endpoints. Large Language Models have shown to be capable of automatically creating system integrations (e.g., as service composition) based on this documentation but require concise input due to input oken limitations, especially regarding comprehensive API descriptions. Currently, it is unknown how best to preprocess these API descriptions. In the present work, we (i) analyze the usage of Retrieval Augmented Generation for endpoint discovery and the chunking, i.e., preprocessing, of state-of-practice OpenAPIs to reduce the input oken length while preserving the most relevant information. To further reduce the input token length for the composition prompt and improve endpoint retrieval, we propose (ii) a Discovery Agent that only receives a summary of the most relevant endpoints nd retrieves specification details on demand. We evaluate RAG for endpoint discovery using (iii) a proposed novel service discovery benchmark SOCBench-D representing a general setting across numerous domains and the real-world RestBench enchmark, first, for the different chunking possibilities and parameters measuring the endpoint retrieval accuracy. Then, we assess the Discovery Agent using the same test data set. The prototype shows how to successfully employ RAG for endpoint discovery to reduce the token count. Our experiments show that endpoint-based approaches outperform naive chunking methods for preprocessing. Relying on an agent significantly improves precision while being prone to decrease recall, disclosing the need for further reasoning capabilities.

Recent advances in large language models (LLMs) demonstrate their impressive reasoning capabilities. However, the reasoning confined to internal parametric space limits LLMs' access to real-time information and understanding of the physical world. To overcome this constraint, we introduce SituatedThinker, a novel framework that enables LLMs to ground their reasoning in real-world contexts through situated thinking, which adaptively combines both internal knowledge and external information with predefined interfaces. By utilizing reinforcement learning, SituatedThinker incentivizes deliberate reasoning with the real world to acquire information and feedback, allowing LLMs to surpass their knowledge boundaries and enhance reasoning. Experimental results demonstrate significant performance improvements on multi-hop question-answering and mathematical reasoning benchmarks. Furthermore, SituatedThinker demonstrates strong performance on unseen tasks, such as KBQA, TableQA, and text-based games, showcasing the generalizable real-world grounded reasoning capability. Our codes are available at https://github.com/jnanliu/SituatedThinker.

Recent advancements in large audio language models (LALMs) have demonstrated impressive results and promising prospects in universal understanding and reasoning across speech, music, and general sound. However, these models still lack the ability to recognize their knowledge boundaries and refuse to answer questions they don't know proactively. While there have been successful attempts to enhance the reliability of LLMs, reliable LALMs remain largely unexplored. In this paper, we systematically investigate various approaches towards reliable LALMs, including training-free methods such as multi-modal chain-of-thought (MCoT), and training-based methods such as supervised fine-tuning (SFT). Besides, we identify the limitations of previous evaluation metrics and propose a new metric, the Reliability Gain Index (RGI), to assess the effectiveness of different reliable methods. Our findings suggest that both training-free and training-based methods enhance the reliability of LALMs to different extents. Moreover, we find that awareness of reliability is a "meta ability", which can be transferred across different audio modalities, although significant structural and content differences exist among sound, music, and speech.