Warning: This paper contains inappropriate/harmful visual contents. While widespread access to the Internet and the rapid advancement of generative models boost people's creativity and productivity, the risk of encountering inappropriate or harmful content also increases.

Large-scale pre-training and instruction tuning have been successful at creating general-purpose language models with broad competence. However, building general-purpose vision-language models is challenging due to the rich input distributions and task diversity resulting from the additional visual input. Although vision-language pretraining has been widely studied, vision-language instruction tuning remains under-explored. In this paper, we conduct a systematic and comprehensive study on vision-language instruction tuning based on the pretrained BLIP-2 models. We gather 26 publicly available datasets, covering a wide variety of tasks and capabilities, and transform them into instruction tuning format.

Abstract--Large Vision-Language Models (VLMs) often exhibit text inertia, where attention drifts from visual evidence toward linguistic priors, resulting in object hallucinations. Existing decoding strategies intervene only at the output logits and thus cannot correct internal reasoning drift, while recent internal-control methods based on heuristic head suppression or global steering vectors lack principled grounding. We introduce Conscious Gaze (CG-VLM), a training-free, inference-time framework that converts game-theoretic interpretability into actionable decoding control. A Cognitive Demand Sensor built on Harsanyi interactions estimates instantaneous vision-text synergy and identifies moments when visual grounding is necessary. CG-VLM achieves state-of-the-art results on POPE and CHAIR across InstructBLIP, LLaV A, Qwen-VL, and mPLUG, while preserving general capabilities, demonstrating that token-level sensing enables precise, context-aware intervention without compromising foundational knowledge.

News image captioning aims to produce journalistically informative descriptions by combining visual content with contextual cues from associated articles. Despite recent advances, existing methods struggle with three key challenges: (1) incomplete information coverage, (2) weak cross-modal alignment, and (3) suboptimal visual-entity grounding. To address these issues, we introduce MERGE, the first Multimodal Entity-aware Retrieval-augmented GEneration framework for news image captioning. MERGE constructs an entity-centric multimodal knowledge base (EMKB) that integrates textual, visual, and structured knowledge, enabling enriched background retrieval. It improves cross-modal alignment through a multistage hypothesis-caption strategy and enhances visual-entity matching via dynamic retrieval guided by image content. Extensive experiments on GoodNews and NYTimes800k show that MERGE significantly outperforms state-of-the-art baselines, with CIDEr gains of +6.84 and +1.16 in caption quality, and F1-score improvements of +4.14 and +2.64 in named entity recognition. Notably, MERGE also generalizes well to the unseen Visual News dataset, achieving +20.17 in CIDEr and +6.22 in F1-score, demonstrating strong robustness and domain adaptability.

Large Vision-Language Models (L VLMs) often suffer from object hallucination, generating text inconsistent with visual inputs, which can critically undermine their reliability. Existing inference-time interventions to mitigate this issue present a challenging trade-off: while methods that steer internal states or adjust output logits can be effective, they often incur substantial computational overhead, typically requiring extra forward passes. This efficiency bottleneck can limit their practicality for real-world, latency-sensitive deployments. In this work, we aim to address this trade-off with Residual-Update Directed DEcoding Regulation (RUDDER), a low-overhead framework that steers L VLMs towards visually-grounded generation. RUDDER is built on two key innovations: (1) Contextual Activation Residual Direction (CARD) vector, a per-sample visual evidence vector extracted from the residual update of a self-attention layer during a single, standard forward pass. Extensive experiments on key hallucination benchmarks, including POPE and CHAIR, indicate that RUDDER achieves performance comparable to state-of-the-art methods while introducing negligible computational latency, validating RUDDER as a pragmatic and effective approach for improving L VLMs' reliability without a significant compromise on efficiency. Code is available at https://anonymous.4open.science/r/ While Large Vision-Language Models (L VLMs) have shown remarkable capabilities in multimodal tasks and are increasingly deployed to assist with real-world problems (Alayrac et al., 2022; Liu et al., 2024a), their practical reliability is critically undermined by a persistent challenge: object hallucination. As shown in Figure 1, L VLMs frequently generate fluent, convincing text that is factually inconsistent with visual groundings, severely limiting their real-world utility and credibility (Ji et al., 2023).

Vision-Language Models (VLMs) have garnered significant attention for their remarkable ability to interpret and generate multimodal content. However, securing these models against jailbreak attacks continues to be a substantial challenge. Unlike text-only models, VLMs integrate additional modalities, introducing novel vulnerabilities such as image hijacking, which can manipulate the model into producing inappropriate or harmful responses. Drawing inspiration from text-based jailbreaks like the "Do Anything Now" (DAN) command, this work introduces VisualDAN, a single adversarial image embedded with DAN-style commands. Specifically, we prepend harmful corpora with affirmative prefixes (e.g., "Sure, I can provide the guidance you need") to trick the model into responding positively to malicious queries. The adversarial image is then trained on these DAN-inspired harmful texts and transformed into the text domain to elicit malicious outputs. Extensive experiments on models such as MiniGPT-4, MiniGPT-v2, InstructBLIP, and LLaVA reveal that VisualDAN effectively bypasses the safeguards of aligned VLMs, forcing them to execute a broad range of harmful instructions that severely violate ethical standards. Our results further demonstrate that even a small amount of toxic content can significantly amplify harmful outputs once the model's defenses are compromised. These findings highlight the urgent need for robust defenses against image-based attacks and offer critical insights for future research into the alignment and security of VLMs.

Large Vision-Language Models (LVLMs) have shown strong performance across multimodal tasks. However, they often produce hallucinations -- text that is inconsistent with visual input, due to the limited ability to verify information in different regions of the image. To address this, we propose Multi-Region Fusion Decoding (MRFD), a training-free decoding method that improves factual grounding by modeling inter-region consistency. MRFD identifies salient regions using cross-attention, generates initial responses for each, and computes reliability weights based on Jensen-Shannon Divergence (JSD) among the responses. These weights guide a consistency-aware fusion of per-region predictions, using region-aware prompts inspired by Chain-of-Thought reasoning. Experiments across multiple LVLMs and benchmarks show that MRFD significantly reduces hallucinations and improves response factuality without requiring model updates.

Warning: This paper contains inappropriate/harmful visual contents. While widespread access to the Internet and the rapid advancement of generative models boost people's creativity and productivity, the risk of encountering inappropriate or harmful content also increases.

Large-scale pre-training and instruction tuning have been successful at creating general-purpose language models with broad competence. However, building general-purpose vision-language models is challenging due to the rich input distributions and task diversity resulting from the additional visual input.