Recent advances have shown that multimodal large language models (MLLMs) benefit from multimodal interleaved chain-of-thought (CoT) with vision tool interactions. However, existing open-source models often exhibit blind tool-use reasoning patterns, invoking vision tools even when they are unnecessary, which significantly increases inference overhead and degrades model performance. To this end, we propose AdaTooler-V, an MLLM that performs adaptive tool-use by determining whether a visual problem truly requires tools. First, we introduce AT-GRPO, a reinforcement learning algorithm that adaptively adjusts reward scales based on the Tool Benefit Score of each sample, encouraging the model to invoke tools only when they provide genuine improvements. Moreover, we construct two datasets to support training: AdaTooler-V-CoT-100k for SFT cold start and AdaTooler-V-300k for RL with verifiable rewards across single-image, multi-image, and video data. Experiments across twelve benchmarks demonstrate the strong reasoning capability of AdaTooler-V, outperforming existing methods in diverse visual reasoning tasks. Notably, AdaTooler-V-7B achieves an accuracy of 89.8\% on the high-resolution benchmark V*, surpassing the commercial proprietary model GPT-4o and Gemini 1.5 Pro. All code, models, and data are released.
Adverse childhood experiences (ACEs) contribute to 75% of mental illness cases in the UK before age 24, yet their emotional impacts are rarely explored through young people's perspectives. This study investigates how youth with ACEs use blurriness in photography as a form of emotional expression and narrative control. Using a participatory methodology, young people acted as co-researchers through photography tasks, Jamboard discussions and blog reflections. Blurred images - emerging spontaneously - became key artefacts for reflection and meaning-making. Thematic analysis, informed by Constructivist and Chaos Theories, found that blurriness symbolised confusion, fragmentation, vulnerability and distance. It also offered a way to express difficult emotions while avoiding overexposure. Participants associated blurred photography with youth visual culture, especially social media aesthetics that value imperfection and authenticity. This research demonstrates the potential of arts-based, co-produced methods to amplify marginalised youth voices and proposes a participatory visual framework for exploring emotional expression among young people affected by ACEs.
Diffusion-based large language models (dLLMs) are trained to model extreme flexibility/dependence in the data-distribution; however, how to best utilize this at inference time remains an open problem. In this work, we uncover an interesting property of these models: dLLMs {trained on textual data} implicitly learn a mixture of semi-autoregressive experts, where different generation orders reveal different specialized behaviors. We show that committing to any single, fixed inference time schedule, a common practice, collapses performance by failing to leverage this latent ensemble. To address this, we introduce HEX (Hidden semi-autoregressive EXperts for test-time scaling), a training-free inference method that ensembles across heterogeneous block schedules. By doing a majority vote over diverse block-sized generation paths, HEX robustly avoids failure modes associated with any single fixed schedule. On reasoning benchmarks such as GSM8K, it boosts accuracy by up to 3.56× (from 24.72\% to 88.10\%), outperforming top-K margin inference and specialized fine-tuned methods like GRPO, without additional training. HEX even yields significant gains on MATH benchmark from 16.40\% to 40.00\%, scientific reasoning on ARC-C from 54.18\% to 87.80\%, and TruthfulQA from 28.36\% to 57.46\%. Our results establish test-time scaling as a powerful principle for dLLMs, showing that the sequence in which masking is done can play a significant role in test-time scaling/inferencing of dLLMs.
Detecting disinformation that blends manipulated text and images has become increasingly challenging, as AI tools make synthetic content easy to generate and disseminate. While most existing AI-safety benchmarks focus on single-modality misinformation (i.e., false content shared without intent to deceive), intentional multimodal disinformation, such as propaganda or conspiracy theories that imitate credible news; remains largely unaddressed. In this work, we introduce the Vision-Language Disinformation Detection Benchmark (VLDBench), the first large-scale resource supporting both unimodal (text-only) and multimodal (text + image) disinformation detection. VLDBench comprises approximately 62,000 labeled text-image pairs across 13 categories, curated from 58 news outlets. Using a semi-automated pipeline followed by expert review, 22 domain experts invested over 500 hours to produce high-quality annotations with substantial inter-annotator agreement. Evaluation of state-of-the-art LLMs and VLMs on VLDBench shows that adding visual cues improves detection accuracy, with gains ranging from 5 points for strong baselines (e.g., LLaMA-3.2-11B-Vision 74.82% vs. LLaMA-3.2-1B-Instruct 70.29%) to 25-30 points for smaller families (e.g., LLaVA-v1.5-Vicuna7B 72.32% vs. Vicuna-7B-v1.5 55.21%), reflecting complementary evidence from images (e.g., meme-like visuals, image-text consistency) that text alone cannot capture. We provide data and code for evaluation, fine-tuning and robustness tests to support disinformation analysis. Developed in alignment with the AI Goverance frameworks (MIT AI Risk Repository), VLDBench offers a principled foundation for advancing trustworthy disinformation detection in multimodal media.
Mixture-of-experts variants of parameter-efficient fine-tuning enable per-token specialization, but they introduce additional trainable routers and expert parameters, increasing memory usage and training cost. This undermines the core goal of parameter-efficient fine-tuning. We propose Monkey Jump, a method that brings mixture-of-experts-style specialization to parameter-efficient fine-tuning without introducing extra trainable parameters for experts or routers. Instead of adding new adapters as experts, Monkey Jump treats the adapters already present in each Transformer block (such as query, key, value, up, and down projections) as implicit experts and routes tokens among them. Routing is performed using k-means clustering with exponentially moving averaged cluster centers, requiring no gradients and no learned parameters. We theoretically show that token-wise routing increases expressivity and can outperform shared adapters by avoiding cancellation effects. Across multi-task experiments covering 14 text, 14 image, and 19 video benchmarks, Monkey Jump achieves competitive performance with mixture-of-experts-based parameter-efficient fine-tuning methods while using 7 to 29 times fewer trainable parameters, up to 48 percent lower memory consumption, and 1.5 to 2 times faster training. Monkey Jump is architecture-agnostic and can be applied to any adapter-based parameter-efficient fine-tuning method.
