Video comprises the vast majority of bits that are generated daily, and is the primary signal driving current innovations in robotics, remote sensing, and wearable technology. Yet, the most powerful video understanding models are too expensive for the resource-constrained platforms used in these applications. One approach is to offload inference to the cloud; this gives access to GPUs capable of processing high-resolution videos in real time. But even with reliable, high-bandwidth communication channels, the combined latency of video encoding, model inference, and round-trip communication prohibits use for certain real-time applications. The alternative is to use fully local inference; but this places extreme constraints on computational and power costs, requiring smaller models and lower resolution, leading to degraded accuracy. To address these challenges, we propose Dedelayed, a real-time inference system that divides computation between a remote model operating on delayed video frames and a local model with access to the current frame. The remote model is trained to make predictions on anticipated future frames, which the local model incorporates into its prediction for the current frame. The local and remote models are jointly optimized with an autoencoder that limits the transmission bitrate required by the available downlink communication channel. We evaluate Dedelayed on the task of real-time streaming video segmentation using the BDD100k driving dataset. For a round trip delay of 100 ms, Dedelayed improves performance by 6.4 mIoU compared to fully local inference and 9.8 mIoU compared to remote inference -- an equivalent improvement to using a model ten times larger.

Large language models (LLMs) have shown promising accuracy in predicting survey responses and policy preferences, which has increased interest in their potential to represent human interests in various domains. Most existing research has focused on "behavioral cloning", effectively evaluating how well models reproduce individuals' expressed preferences. Drawing on theories of political representation, we highlight an underexplored design trade-off: whether AI systems should act as delegates, mirroring expressed preferences, or as trustees, exercising judgment about what best serves an individual's interests. This trade-off is closely related to issues of LLM sycophancy, where models can encourage behavior or validate beliefs that may be aligned with a user's short-term preferences, but is detrimental to their long-term interests. Through a series of experiments simulating votes on various policy issues in the U.S. context, we apply a temporal utility framework that weighs short and long-term interests (simulating a trustee role) and compare voting outcomes to behavior-cloning models (simulating a delegate). We find that trustee-style predictions weighted toward long-term interests produce policy decisions that align more closely with expert consensus on well-understood issues, but also show greater bias toward models' default stances on topics lacking clear agreement. These findings reveal a fundamental trade-off in designing AI systems to represent human interests. Delegate models better preserve user autonomy but may diverge from well-supported policy positions, while trustee models can promote welfare on well-understood issues yet risk paternalism and bias on subjective topics.

Reliable handling of code diffs is central to agents that edit and refactor repositories at scale. We introduce Diff-XYZ, a compact benchmark for code-diff understanding with three supervised tasks: apply (old code $+$ diff $\rightarrow$ new code), anti-apply (new code $-$ diff $\rightarrow$ old code), and diff generation (new code $-$ old code $\rightarrow$ diff). Instances in the benchmark are triples $\langle \textit{old code}, \textit{new code}, \textit{diff} \rangle$ drawn from real commits in CommitPackFT, paired with automatic metrics and a clear evaluation protocol. We use the benchmark to do a focused empirical study of the unified diff format and run a cross-format comparison of different diff representations. Our findings reveal that different formats should be used depending on the use case and model size. For example, representing diffs in search-replace format performs best for larger models across most tasks, while structured udiff variants offer similar but slightly weaker performance. In contrast, smaller open models benefit little from any formatting choice. The Diff-XYZ benchmark is a reusable foundation for assessing and improving diff handling in LLMs that can aid future development of diff formats and models editing code. The dataset is published on HuggingFace Hub: https://huggingface.co/datasets/JetBrains-Research/diff-xyz.

Knowledge-based Visual Question Answering (KVQA) requires models to ground entities in images and reason over factual knowledge. Recent work has introduced its implicit-knowledge variant, IK-KVQA, where a multimodal large language model (MLLM) is the sole knowledge source and answers are produced without external retrieval. Existing IK-KVQA approaches, however, are typically trained with answer-only supervision: reasoning remains implicit, justifications are often weak or inconsistent, and generalization after standard supervised fine-tuning (SFT) can be brittle. We propose MODELNAME, a framework that equips IK-KVQA with dual-path structured reasoning traces (symbolic relation paths over text and vision together with path-grounded natural-language explanations) to provide a stronger inductive bias than generic answer-only supervision. These traces act as modality-aware scaffolds that guide the model toward relevant entities and attributes, offering more structure than generic chain-of-thought supervision while not constraining reasoning to any single fixed path. Using a single open-source MLLM, MODELNAME constructs and selects traces to build an offline trace-enriched dataset and then performs structure-aware self-distillation; no external retrievers, verifiers, or curated knowledge bases are used, and inference is a single autoregressive pass. Across benchmarks, MODELNAME consistently improves both answer accuracy and the transparency of intermediate reasoning, achieving up to 11.3% higher answer accuracy on OK-VQA over the strongest baseline.

Diffusion language models have recently emerged as a competitive alternative to autoregressive language models. Beyond next-token generation, they are more efficient and flexible by enabling parallel and any-order token generation. However, despite empirical successes, their computational power and fundamental limitations remain poorly understood. In this paper, we formally study whether non-autoregressive generation in Masked Diffusion Models (MDM) enables solving problems beyond the reach of Auto-Regressive Models (ARM). Our results show that MDM with sufficiently large context length is computationally universal with decoding steps matching the optimal parallel time complexity in PRAM. However, when controlling for other factors, MDM's flexibility to generate in any-order does not expand what ARM can already solve. To address this, we propose a new form of generation called any-process generation, which extends MDM with capabilities to remask, insert and delete tokens, allowing self-correction, length-variable editing, and adaptive parallelism. Theoretically and empirically, we demonstrate these capabilities enable scalability to significantly harder reasoning problems that are otherwise intractable for ARM and vanilla MDM. Additionally, they prove essential for generation tasks where objects naturally evolve through non-sequential processes, crucial for extending current LLMs beyond natural language to domains such as coding and science.

CAPTCHA, originally designed to distinguish humans from robots, has evolved into a real-world benchmark for assessing the spatial reasoning capabilities of vision-language models. In this work, we first show that step-by-step reasoning is crucial for vision-language models (VLMs) to solve CAPTCHAs, which represent high-difficulty spatial reasoning tasks, and that current commercial vision-language models still struggle with such reasoning. In particular, we observe that most commercial VLMs (e.g., Gemini, Claude, GPT, etc.) fail to effectively solve CAPTCHAs and thus achieve low accuracy (around 21.9 percent). However, our findings indicate that requiring the model to perform step-by-step reasoning before generating the final coordinates can significantly enhance its solving accuracy, underscoring the severity of the gap. To systematically study this issue, we introduce CAPTCHA-X, the first real-world CAPTCHA benchmark with reasoning, covering seven categories of CAPTCHAs (such as Gobang, hCaptcha, etc.) with step-by-step action solutions and grounding annotations. We further define five reasoning-oriented metrics that enable a comprehensive evaluation of models reasoning capabilities. To validate the effectiveness of reasoning, we also propose a general agentic VLM-based framework that incorporates the models inherent reasoning abilities. Our method achieves state-of-the-art performance across five high-difficulty CAPTCHA types, with an average solving accuracy of 83.9 percent, substantially surpassing existing baselines. These results reveal the limitations of current models and highlight the importance of reasoning in advancing visual-spatial challenges in the future.

Motivated by the remarkable success of Foundation Models (FMs) in language modeling, there has been growing interest in developing FMs for time series prediction, given the transformative power such models hold for science and engineering. This culminated in significant success of FMs in short-range forecasting settings. However, extrapolation or long-range forecasting remains elusive for FMs, which struggle to outperform even simple baselines. This contrasts with physical laws which have strong extrapolation properties, and raises the question of the fundamental difference between the structure of neural networks and physical laws. In this work, we identify and formalize a fundamental property characterizing the ability of statistical learning models to predict more accurately outside of their training domain, hence explaining performance deterioration for deep learning models in extrapolation settings. In addition to a theoretical analysis, we present empirical results showcasing the implications of this property on current deep learning architectures. Our results not only clarify the root causes of the extrapolation gap but also suggest directions for designing next-generation forecasting models capable of mastering extrapolation.

Accurate brain tumor classification is critical for intra-operative decision making in neuro-oncological surgery. However, existing approaches are restricted to a fixed set of predefined classes and are therefore unable to capture patterns of tumor types not available during training. Unsupervised learning can extract general-purpose features, but it lacks the ability to incorporate prior knowledge from labelled data, and semi-supervised methods often assume that all potential classes are represented in the labelled data. Generalized Category Discovery (GCD) aims to bridge this gap by categorizing both known and unknown classes within unlabelled data. To reflect the hierarchical structure of brain tumor taxonomies, in this work, we introduce Hierarchical Generalized Category Discovery for Brain Tumor Classification (HGCD-BT), a novel approach that integrates hierarchical clustering with contrastive learning. Our method extends contrastive learning based GCD by incorporating a novel semi-supervised hierarchical clustering loss. We evaluate HGCD-BT on OpenSRH, a dataset of stimulated Raman histology brain tumor images, achieving a +28% improvement in accuracy over state-of-the-art GCD methods for patch-level classification, particularly in identifying previously unseen tumor categories. Furthermore, we demonstrate the generalizability of HGCD-BT on slide-level classification of hematoxylin and eosin stained whole-slide images from the Digital Brain Tumor Atlas, confirming its utility across imaging modalities.

Large Language Models (LLMs) have demonstrated remarkable capabilities in various tasks. However, they remain exposed to jailbreak attacks, eliciting harmful responses. The nested scenario strategy has been increasingly adopted across various methods, demonstrating immense potential. Nevertheless, these methods are easily detectable due to their prominent malicious intentions. In this work, we are the first to find and systematically verify that LLMs' alignment defenses are not sensitive to nested scenarios, where these scenarios are highly semantically relevant to the queries and incorporate targeted toxic knowledge. This is a crucial yet insufficiently explored direction. Based on this, we propose RTS-Attack (Semantically Relevant Nested Scenarios with Targeted Toxic Knowledge), an adaptive and automated framework to examine LLMs' alignment. By building scenarios highly relevant to the queries and integrating targeted toxic knowledge, RTS-Attack bypasses the alignment defenses of LLMs. Moreover, the jailbreak prompts generated by RTS-Attack are free from harmful queries, leading to outstanding concealment. Extensive experiments demonstrate that RTS-Attack exhibits superior performance in both efficiency and universality compared to the baselines across diverse advanced LLMs, including GPT-4o, Llama3-70b, and Gemini-pro. Our complete code is available at https://github.com/nercode/Work. WARNING: THIS PAPER CONTAINS POTENTIALLY HARMFUL CONTENT.

REtrieval-Augmented LLM-based Machine Translation (REAL-MT) shows promise for knowledge-intensive tasks like idiomatic translation, but its reliability under noisy retrieval, a common challenge in real-world deployment, remains poorly understood. To address this gap, we propose a noise synthesis framework and new metrics to systematically evaluate REAL-MT's reliability across high-, medium-, and low-resource language pairs. Using both open-and closed-sourced models, including standard LLMs and large reasoning models (LRMs), we find that models heavily rely on retrieved context, and this dependence is significantly more detrimental in low-resource language pairs, producing nonsensical translations. Although LRMs possess enhanced reasoning capabilities, they show no improvement in error correction and are even more susceptible to noise, tending to rationalize incorrect contexts. Attention analysis reveals a shift from the source idiom to noisy content, while confidence increases despite declining accuracy, indicating poor self-monitoring. To mitigate these issues, we investigate training-free and fine-tuning strategies, which improve robustness at the cost of performance in clean contexts, revealing a fundamental trade-off. Our findings highlight the limitations of current approaches, underscoring the need for self-verifying integration mechanisms.