搜索结果：Happened

Existing training-free approaches for GUI grounding often rely on multiple inference runs, such as iterative cropping or candidate aggregation, to identify target elements. Despite this additional computation, each forward pass still independently interprets the instruction and parses the visual layout, without enabling progressive interaction among visual tokens. In this paper, we study what happens during GUI grounding in Vision-Language Models (VLMs) and identify a previously overlooked bottleneck. We show that grounding follows a two-stage paradigm: the prefill stage determines candidate UI elements, while the decoding stage subsequently refines the final coordinates. This asymmetry establishes prefill as the critical step, as errors in candidate selection cannot be effectively corrected during decoding. Based on this observation, we propose Re-Prefill, a training-free method that revisits inference by introducing an attention-guided second prefill stage to refine target selection. Specifically, visual tokens that consistently receive high attention from the query position, i.e., the final token, across layers are extracted as a preliminary target hypothesis and appended to the

Traffic safety science has long been hindered by a fundamental data paradox: the crashes we most wish to prevent are precisely those events we rarely observe. Existing crash-frequency models and surrogate safety metrics rely heavily on sparse, noisy, and under-reported records, while even sophisticated, high-fidelity simulations undersample the long-tailed situations that trigger catastrophic outcomes such as fatalities. We argue that the path to achieving Vision Zero, i.e., the complete elimination of traffic fatalities and severe injuries, requires a paradigm shift from traditional crash-only learning to a new form of counterfactual safety learning: reasoning not only about what happened, but also about the vast set of plausible yet perilous scenarios that could have happened under slightly different circumstances. To operationalize this shift, our proposed agenda bridges macro to micro. Guided by crash-rate priors, generative scene engines, diverse driver models, and causal learning, near-miss events are synthesized and explained. A crash-focused digital twin testbed links micro scenes to macro patterns, while a multi-objective validator ensures that simulations maintain statist

The claim that at the so-called Planck scale our current physics breaks down and a new theory of quantum gravity is required is ubiquitous, but the evidence is shakier than the confidence of those assertions warrants. In this paper, I survey five arguments in favour of this claim - based on dimensional analysis, quantum black holes, generalised uncertainty principles, the nonrenormalisability of quantum gravity, and theories beyond the standard model - but find that none of them succeeds. The argument from nonrenormalisability is the most convincing, yet it requires the unwarranted assumption that the same constant of action occurs in every quantum field theory. Therefore, our theories don't (yet) predict that quantum gravity happens at the Planck scale.

Autonomous Vehicle decisions rely on multimodal prediction models that account for multiple route options and the inherent uncertainty in human behavior. However, models can suffer from mode collapse, where only the most likely mode is predicted, posing significant safety risks. While existing methods employ various strategies to generate diverse predictions, they often overlook the diversity in interaction modes among agents. Additionally, traditional metrics for evaluating prediction models are dataset-dependent and do not evaluate inter-agent interactions quantitatively. To our knowledge, none of the existing metrics explicitly evaluates mode collapse. In this paper, we propose a novel evaluation framework that assesses mode collapse in joint trajectory predictions, focusing on safety-critical interactions. We introduce metrics for mode collapse, mode correctness, and coverage, emphasizing the sequential dimension of predictions. By testing four multi-agent trajectory prediction models, we demonstrate that mode collapse indeed happens. When looking at the sequential dimension, although prediction accuracy improves closer to interaction events, there are still cases where the mod

Recent unified models for joint understanding and generation have significantly advanced visual generation capabilities. However, their focus on conventional tasks like text-to-video generation has left the temporal reasoning potential of unified models largely underexplored. To address this gap, we introduce Next Scene Prediction (NSP), a new task that pushes unified video models toward temporal and causal reasoning. Unlike text-to-video generation, NSP requires predicting plausible futures from preceding context, demanding deeper understanding and reasoning. To tackle this task, we propose a unified framework combining Qwen-VL for comprehension and LTX for synthesis, bridged by a latent query embedding and a connector module. This model is trained in three stages on our newly curated, large-scale NSP dataset: text-to-video pre-training, supervised fine-tuning, and reinforcement learning (via GRPO) with our proposed causal consistency reward. Experiments demonstrate our model achieves state-of-the-art performance on our benchmark, advancing the capability of generalist multimodal systems to anticipate what happens next.

We initiate a unified theoretical and algorithmic study of a key problem in weak-to-strong (W2S) generalization: when fine-tuning a strong pre-trained student with pseudolabels from a weaker teacher on a downstream task with spurious correlations, does W2S happen, and how to improve it upon failures? We consider two sources of spurious correlations caused by group imbalance: (i) a weak teacher fine-tuned on group-imbalanced labeled data with a minority group of fraction $η_\ell$, and (ii) a group-imbalanced unlabeled set pseudolabeled by the teacher with a minority group of fraction $η_u$. Theoretically, a precise characterization of W2S gain at the proportional asymptotic limit shows that W2S always happens with sufficient pseudolabels when $η_u = η_\ell$ but may fail when $η_u e η_\ell$, where W2S gain diminishes as $(η_u - η_\ell)^2$ increases. Our theory is corroborated by extensive experiments on various spurious correlation benchmarks and teacher-student pairs. To boost W2S performance upon failures, we further propose a simple, effective algorithmic remedy that retrains the strong student on its high-confidence data subset after W2S fine-tuning. Our algorithm is group-label

Effective demand forecasting is critical for inventory management, production planning, and decision making across industries. Selecting the appropriate model and suitable features to efficiently capture patterns in the data is one of the main challenges in demand forecasting. In reality, this becomes even more complicated when the recorded sales have zeroes, which can happen naturally or due to some anomalies, such as stockouts and recording errors. Mistreating the zeroes can lead to the application of inappropriate forecasting methods, and thus leading to poor decision making. Furthermore, the demand itself can have different fundamental characteristics, and being able to distinguish one type from another might bring substantial benefits in terms of accuracy and thus decision making. We propose a two-stage model-based classification framework that in the first step, identifies artificially occurring zeroes, and in the second, classifies demand to one of the possible types: regular/intermittent, intermittent smooth/lumpy, fractional/count. The framework relies on statistical modelling and information criteria. We argue that different types of demand need different features, and sh

Prediction markets are starting to look less like crowd polls and more like electronic markets. The central question is therefore no longer only whether these markets forecast well, but what happens when institutional liquidity enters: do spreads tighten, does price discovery improve, and do those gains actually reach the traders who are slowest to react when information arrives? This paper offers a research design for answering that question. It defines a broad market-quality lens, separates the main channels through which institutional liquidity enters, and maps the identification problems that arise in live venue data. It also uses a synthetic microstructure laboratory as a proof of concept for the measurement pipeline. The main lesson of the synthetic exercise is deliberately narrow. Market-maker coverage, liquidity incentives, and automation do not have to work through the same channel; average liquidity gains do not have to translate into equal gains for all traders; and the sharpest welfare losses are most likely to appear in shock states, when slower takers receive the least pass-through of tighter quoted markets. The synthetic results are useful because they stress-test th

The internet serves as a common source of training data for generative AI (genAI) models but is increasingly populated with AI-generated content. This duality raises the possibility that future genAI models may be trained on other models' generated outputs. Prior work has studied consequences of models training on their own generated outputs, but limited work has considered what happens if models ingest content produced by other models. Given society's increasing dependence on genAI tools, understanding such data-mediated model interactions is critical. This work provides empirical evidence for how data-mediated interactions might unfold in practice, develops a theoretical model for this interactive training process, and experimentally validates the theory. We find that data-mediated interactions can benefit models by exposing them to novel concepts perhaps missed in original training data, but also can homogenize their performance on shared tasks.

We study how the rigidity transition in a triangular lattice changes as a function of anisotropy by preferentially filling bonds on the lattice in one direction. We discover that the onset of rigidity in anisotropic spring networks arises in at least two steps, reminiscent of the two-step melting transition in two dimensional crystals. In particular, our simulations demonstrate that the percolation of stress-supporting bonds happens at different critical volume fractions along different directions. By examining each independent component of the elasticity tensor, we determine universal exponents and develop universal scaling functions to analyze isotropic rigidity percolation as a multicritical point. We expect that these results will be important for elucidating the underlying mechanical phase transitions governing the properties of biological materials ranging from the cytoskeletons of cells to the extracellular networks of tissues such as tendon where the networks are often preferentially aligned.

We investigate the behavior of methods that use linear projections to remove information about a concept from a language representation, and we consider the question of what happens to a dataset transformed by such a method. A theoretical analysis and experiments on real-world and synthetic data show that these methods inject strong statistical dependencies into the transformed datasets. After applying such a method, the representation space is highly structured: in the transformed space, an instance tends to be located near instances of the opposite label. As a consequence, the original labeling can in some cases be reconstructed by applying an anti-clustering method.