We introduce DRISHTIKON, a first-of-its-kind multimodal and multilingual benchmark centered exclusively on Indian culture, designed to evaluate the cultural understanding of generative AI systems. Unlike existing benchmarks with a generic or global scope, DRISHTIKON offers deep, fine-grained coverage across India's diverse regions, spanning 15 languages, covering all states and union territories, and incorporating over 64,000 aligned text-image pairs. The dataset captures rich cultural themes including festivals, attire, cuisines, art forms, and historical heritage amongst many more. We evaluate a wide range of vision-language models (VLMs), including open-source small and large models, proprietary systems, reasoning-specialized VLMs, and Indic-focused models, across zero-shot and chain-of-thought settings. Our results expose key limitations in current models' ability to reason over culturally grounded, multimodal inputs, particularly for low-resource languages and less-documented traditions. DRISHTIKON fills a vital gap in inclusive AI research, offering a robust testbed to advance culturally aware, multimodally competent language technologies.

Automatic Speech Recognition (ASR) for low-resource languages like Slovak is hindered by the scarcity of training data. To address this, we introduce SloPalSpeech, a new, large-scale Slovak ASR dataset containing 2,806 hours of speech from parliamentary proceedings. We developed a robust processing pipeline to align and segment long-form recordings into clean, 30-second audio-transcript pairs suitable for model training. We use this dataset to fine-tune several OpenAI Whisper models (small, medium, large-v3, and large-v3-turbo), achieving significant Word Error Rate (WER) reductions on standard Slovak benchmarks like Common Voice and FLEURS. For instance, the fine-tuned Whisper-small model's WER dropped by up to 70\%, approaching the baseline performance of the much larger Whisper-large-v3 model. To foster future research in low-resource speech recognition, we publicly release the complete SloPalSpeech dataset, the fully segmented transcripts (60 million words), and all our fine-tuned models.

Abstract--Continuous human motion understanding remains a core challenge in computer vision due to its high dimensionality and inherent redundancy. Efficient compression and representation are crucial for analyzing complex motion dynamics. In this work, we introduce an adversarially-refined VQ-GAN framework with dense motion tokenization for compressing spatio-temporal heatmaps while preserving the fine-grained traces of human motion. Our approach combines dense motion tokenization with adversarial refinement, which eliminates reconstruction artifacts like motion smearing and temporal misalignment observed in non-adversarial baselines. Our experiments on the CMU Panoptic dataset [7] provide conclusive evidence of our method's superiority, outperforming the dV AE baseline by 9.31% SSIM and reducing temporal instability by 37.1%. Furthermore, our dense tokenization strategy enables a novel analysis of motion complexity, revealing that 2D motion can be optimally represented with a compact 128-token vocabulary, while 3D motion's complexity demands a much larger 1024-token codebook for faithful reconstruction. These results establish practical deployment feasibility across diverse motion analysis applications.

Proper initialization is crucial for any system, particularly in multi-agent systems (MAS), where it plays a pivotal role in determining both the system's efficiency and effectiveness. However, existing MAS initialization methods do not fully account for the collaborative needs of the generated agents in subsequent stages. Inspired by the principles of effective team composition, we propose AgentInit, which aims to optimize the structure of agent teams. Specifically, in addition to multi-round interactions and reflections between agents during agent generation, AgentInit incorporates a Natural Language to Format mechanism to ensure consistency and standardization. Balanced team selection strategies using Pareto principles are subsequently applied to jointly consider agent team diversity and task relevance to promote effective and efficient collaboration and enhance overall system performance. Experiments show that AgentInit consistently outperforms state-of-the-art initialization methods and pre-defined strategies across various frameworks and tasks, achieving an overall performance improvement of up to 1.2 and 1.6, respectively, while also significantly reducing token consumption. Further analysis confirms its strong transferability to similar tasks and verifies the effectiveness of its key components, demonstrating its capability and adaptability as a reliable MAS initialization method. Source code and models are available at https://github.com/1737423697/AgentInit.

While soft context compression methods, which map input text to smaller latent representations, have shown promise, their real-world adoption is limited. Existing techniques typically compress the context as a single unit, which leads to quadratic compression complexity and an inability to reuse computations across queries with overlapping contexts. In this work, we introduce CompLLM, a soft compression technique designed for practical deployment. Instead of processing the context holistically, CompLLM divides it into segments and compresses each one independently. This simple design choice yields three critical properties: efficiency, as the compression step scales linearly with the context length; scalability, enabling models trained on short sequences (e.g., 1k tokens) to generalize to contexts of 100k tokens; and reusability, allowing compressed segments to be cached and reused across different queries. Our experiments show that with a 2x compression rate, at high context lengths CompLLM speeds up Time To First Token (TTFT) by up to 4x and reduces the KV cache size by 50%. Furthermore, CompLLM achieves performance comparable to that obtained with the uncompressed context, and even surpasses it on very long sequences, demonstrating its effectiveness and practical utility. LOFT is a long context benchmark (128k tokens) designed to stress-test the long context capabilities of frontiers LLMs as Gemini 1.5 Pro, GPT -4o, and Claude 3 Opus. With CompLLM we show that we can improve long context capabilities of much smaller open source LLMs. Figure 1: At high context lengths, CompLLM leads to considerable speedup and improved results, without requiring any modification or tuning of the LLM, by efficiently reducing the number of embeddings fed to the LLM. The plot shows the Time To First Token (TTFT) with CompLLM and without it (i.e. with a standard pipeline) as a function of context length. Among the many use cases of LLMs, one of the most popular is long context Q&A: given a textual context of arbitrary length, the LLM should answer questions about it. Applications include coding assistants reading large codebases (Team, 2024), web agents reasoning on HTML pages (Zeng et al., 2024), users querying an LLM about a set of documents (Liu et al., 2024a), or RAG systems 1 Due to the quadratic complexity of the transformer (V aswani et al., 2017), processing long contexts can be unfeasibly expensive: it is therefore important to reduce computational complexity, especially as contexts grows longer and longer.

With the widespread deployment of dashcams and advancements in computer vision, developing accident prediction models from the dashcam perspective has become critical for proactive safety interventions. However, two key challenges persist: modeling feature-level interactions among traffic participants (often occluded in dashcam views) and capturing complex, asynchronous multi-temporal behavioral cues preceding accidents. To deal with these two challenges, a Multi-scale Feature Interaction Network (MsFIN) is proposed for early-stage accident anticipation from dashcam videos. MsFIN has three layers for multi-scale feature aggregation, temporal feature processing and multi-scale feature post fusion, respectively. For multi-scale feature aggregation, a Multi-scale Module is designed to extract scene representations at short-term, mid-term and long-term temporal scales. Meanwhile, the Transformer architecture is leveraged to facilitate comprehensive feature interactions. Temporal feature processing captures the sequential evolution of scene and object features under causal constraints. In the multi-scale feature post fusion stage, the network fuses scene and object features across multiple temporal scales to generate a comprehensive risk representation. Experiments on DAD and DADA datasets show that MsFIN significantly outperforms state-of-the-art models with single-scale feature extraction in both prediction correctness and earliness. Ablation studies validate the effectiveness of each module in MsFIN, highlighting how the network achieves superior performance through multi-scale feature fusion and contextual interaction modeling.

Recent advances in text-to-video (T2V) generation have enabled the creation of high-fidelity, temporally coherent clips from natural language prompts. Yet these systems come with significant computational costs, and their energy demands remain poorly understood. In this paper, we present a systematic study of the latency and energy consumption of state-of-the-art open-source T2V models. We first develop a compute-bound analytical model that predicts scaling laws with respect to spatial resolution, temporal length, and denoising steps. We then validate these predictions through fine-grained experiments on WAN2.1-T2V, showing quadratic growth with spatial and temporal dimensions, and linear scaling with the number of denoising steps. Finally, we extend our analysis to six diverse T2V models, comparing their runtime and energy profiles under default settings. Our results provide both a benchmark reference and practical insights for designing and deploying more sustainable generative video systems.

Evaluation of hydrocephalus in children is challenging, and the related research is limited by a lack of publicly available, expert-annotated datasets, particularly those with segmentation of the choroid plexus. To address this, we present HyKid, an open-source dataset from 48 pediatric patients with hydrocephalus. 3D MRIs were provided with 1mm isotropic resolution, which was reconstructed from routine low-resolution images using a slice-to-volume algorithm. Manually corrected segmentations of brain tissues, including white matter, grey matter, lateral ventricle, external CSF, and the choroid plexus, were provided by an experienced neurologist. Additionally, structured data was extracted from clinical radiology reports using a Retrieval-Augmented Generation framework. The strong correlation between choroid plexus volume and total CSF volume provided a potential biomarker for hydrocephalus evaluation, achieving excellent performance in a predictive model (AUC = 0.87). The proposed HyKid dataset provided a high-quality benchmark for neuroimaging algorithms development, and it revealed the choroid plexus-related features in hydrocephalus assessments. Our datasets are publicly available at https://www.synapse.org/Synapse:syn68544889.

Multimodal Large Language Models (MLLMs) are increasingly deployed in real-world applications, yet their ability to make context-aware safety decisions remains limited. Existing methods often fail to balance oversensitivity (unjustified refusals of benign queries) and undersensitivity (missed detection of visually grounded risks), leaving a persistent gap in safety alignment. To address this issue, we introduce Safety-aware Contrastive Decoding (SafeCoDe), a lightweight and model-agnostic decoding framework that dynamically adjusts token generation based on multimodal context. SafeCoDe operates in two stages: (1) a contrastive decoding mechanism that highlights tokens sensitive to visual context by contrasting real and Gaussian-noised images, and (2) a global-aware token modulation strategy that integrates scene-level reasoning with token-level adjustment to adapt refusals according to the predicted safety verdict. Extensive experiments across diverse MLLM architectures and safety benchmarks, covering undersensitivity, oversensitivity, and general safety evaluations, show that SafeCoDe consistently improves context-sensitive refusal behaviors while preserving model helpfulness.

Data-driven models, especially deep learning classifiers often demonstrate great success on clean datasets. Yet, they remain vulnerable to common data distortions such as adversarial and common corruption perturbations. These perturbations can significantly degrade performance, thereby challenging the overall reliability of the models. Traditional robustness validation typically relies on perturbed test datasets to assess and improve model performance. In our framework, however, we propose a validation approach that extracts "weak robust" samples directly from the training dataset via local robustness analysis. These samples, being the most susceptible to perturbations, serve as an early and sensitive indicator of the model's vulnerabilities. By evaluating models on these challenging training instances, we gain a more nuanced understanding of its robustness, which informs targeted performance enhancement. We demonstrate the effectiveness of our approach on models trained with CIFAR-10, CIFAR-100, and ImageNet, highlighting how robustness validation guided by weak robust samples can drive meaningful improvements in model reliability under adversarial and common corruption scenarios.