Tootfinder

SAMFusion: Sensor-Adaptive Multimodal Fusion for 3D Object Detection in Adverse Weather
Edoardo Palladin, Roland Dietze, Praveen Narayanan, Mario Bijelic, Felix Heide
https://arxiv.org/abs/2508.16408

SAMFusion: Sensor-Adaptive Multimodal Fusion for 3D Object Detection in Adverse Weather
Multimodal sensor fusion is an essential capability for autonomous robots, enabling object detection and decision-making in the presence of failing or uncertain inputs. While recent fusion methods excel in normal environmental conditions, these approaches fail in adverse weather, e.g., heavy fog, snow, or obstructions due to soiling. We introduce a novel multi-sensor fusion approach tailored to adverse weather conditions. In addition to fusing RGB and LiDAR sensors, which are employed in recent…

Enhanced Performance of FeFET Gate Stack via Heterogeneously co-doped Ferroelectric HfO$_2$ Films
Shouzhuo Yang, David Lehninger, Peter Reinig, Fred Sch\"one, Raik Hoffmann, Konrad Seidel, Maximilian Lederer, Gerald Gerlach
https://arxiv.org/abs/2508.16768

Enhanced Performance of FeFET Gate Stack via Heterogeneously co-doped Ferroelectric HfO$_2$ Films
In this work, we explore the impact of spatially controlled Zr and Al heterogeneous co-doping in HfO$_2$ thin films tailored for metal-ferroelectric-insulator-semiconductor (MFIS) gate stacks of ferroelectric field effect transistors (FeFETs). By precisely modulating the vertical arrangement of Zr and Al dopants during atomic layer deposition, we introduce deliberate compositional gradients that affect crystallization dynamics during subsequent annealing. This strategy enables us to systematica…

The Conductor and the Engine: A Path Towards Co-Designed Reasoning
Yuanxin Wang, Pawel Filipczuk, Anisha Garg, Amaan Dhada, Mohammad Hassanpour, David Bick, Ganesh Venkatesh
https://arxiv.org/abs/2509.19762

The Conductor and the Engine: A Path Towards Co-Designed Reasoning
Modern LLM reasoning relies on extensive test-time computation, driven by internal model training and external agentic orchestration. However, this synergy is often inefficient, as model verbosity and poor instruction following lead to wasted compute. We analyze this capability-cost trade-off and introduce an optimized reasoning workflow (\cepo) that empowers smaller open-source models to outperform models multiple times their size. We will open-source this workflow to enable further research. …

A notion of quasi-convex subgroups in acylindrically hyperbolic groups
Ping Wan
https://arxiv.org/abs/2509.20532 https://arxiv.org/pdf/2509.20532

A notion of quasi-convex subgroups in acylindrically hyperbolic groups
In this paper, we present a notion of quasiconvexity in the setting of finitely-generated groups with hyperbolically embedded subgroups. Our main result shows that this notion yields uniform quasiconvex constants in the setting of coned-off cusped spaces. We also prove that this notion of quasiconvexity is preserved under sufficiently long Dehn filling. As an application, we generalize a theorem of Groves-Manning on groups acting on $\operatorname{CAT}(0)$ cube complexes.

Interpretable Evaluation of AI-Generated Content with Language-Grounded Sparse Encoders
Yiming Tang, Arash Lagzian, Srinivas Anumasa, Qiran Zou, Trang Nguyen, Ehsan Adeli, Ching-Yu Cheng, Yilun Du, Dianbo Liu
https://arxiv.org/abs/2508.18236

Interpretable Evaluation of AI-Generated Content with Language-Grounded Sparse Encoders
While the quality of AI-generated contents, such as synthetic images, has become remarkably high, current evaluation metrics provide only coarse-grained assessments, failing to identify specific strengths and weaknesses that researchers and practitioners need for model selection and development, further limiting the scientific understanding and commercial deployment of these generative models. To address this, we introduce Language-Grounded Sparse Encoders (LanSE), a novel architecture that cre…

The Conductor and the Engine: A Path Towards Co-Designed Reasoning
Yuanxin Wang, Pawel Filipczuk, Anisha Garg, Amaan Dhada, Mohammad Hassanpour, David Bick, Ganesh Venkatesh
https://arxiv.org/abs/2509.19762

The Conductor and the Engine: A Path Towards Co-Designed Reasoning
Modern LLM reasoning relies on extensive test-time computation, driven by internal model training and external agentic orchestration. However, this synergy is often inefficient, as model verbosity and poor instruction following lead to wasted compute. We analyze this capability-cost trade-off and introduce an optimized reasoning workflow (\cepo) that empowers smaller open-source models to outperform models multiple times their size. We will open-source this workflow to enable further research. …

NewtonGen: Physics-Consistent and Controllable Text-to-Video Generation via Neural Newtonian Dynamics
Yu Yuan, Xijun Wang, Tharindu Wickremasinghe, Zeeshan Nadir, Bole Ma, Stanley H. Chan
https://arxiv.org/abs/2509.21309

NewtonGen: Physics-Consistent and Controllable Text-to-Video Generation via Neural Newtonian Dynamics
A primary bottleneck in large-scale text-to-video generation today is physical consistency and controllability. Despite recent advances, state-of-the-art models often produce unrealistic motions, such as objects falling upward, or abrupt changes in velocity and direction. Moreover, these models lack precise parameter control, struggling to generate physically consistent dynamics under different initial conditions. We argue that this fundamental limitation stems from current models learning moti…

Causal Fuzzing for Verifying Machine Unlearning
Anna Mazhar, Sainyam Galhotra
https://arxiv.org/abs/2509.16525 https://arxiv.org/pdf/2509.16525

Causal Fuzzing for Verifying Machine Unlearning
As machine learning models become increasingly embedded in decision-making systems, the ability to "unlearn" targeted data or features is crucial for enhancing model adaptability, fairness, and privacy in models which involves expensive training. To effectively guide machine unlearning, a thorough testing is essential. Existing methods for verification of machine unlearning provide limited insights, often failing in scenarios where the influence is indirect. In this work, we propose CAFÉ, a ne…

The Unwinnable Arms Race of AI Image Detection
Till Aczel, Lorenzo Vettor, Andreas Plesner, Roger Wattenhofer
https://arxiv.org/abs/2509.21135 https://arxi…

The Unwinnable Arms Race of AI Image Detection
The rapid progress of image generative AI has blurred the boundary between synthetic and real images, fueling an arms race between generators and discriminators. This paper investigates the conditions under which discriminators are most disadvantaged in this competition. We analyze two key factors: data dimensionality and data complexity. While increased dimensionality often strengthens the discriminators ability to detect subtle inconsistencies, complexity introduces a more nuanced effect. Usi…

A New Subclass of Carbon-Enhanced Metal-Poor Stars at Extremely Low Metallicity
Young Sun Lee, Timothy C. Beers, Yutaka Hirai, Jihye Hong, Miji Jeong, Changmin Kim, Young Kwang Kim
https://arxiv.org/abs/2509.16510

A New Subclass of Carbon-Enhanced Metal-Poor Stars at Extremely Low Metallicity
We report the discovery of a new subclass of carbon-enhanced metal-poor (CEMP) stars, characterized by high absolute carbon abundances (A(C) > 7.39) and extremely low metallicity ([Fe/H] $<=$ -3.1) but notably lacking enhancements in neutron-capture elements, thus falling under the CEMP-no category. This population emerged from a detailed analysis of low-resolution spectroscopic data obtained from the Sloan Digital Sky Survey (SDSS) and the Large Sky Area Multi-Object Fiber Spectroscopic Telesc…