Tootfinder

Evaluating the Evaluators: Metrics for Compositional Text-to-Image Generation
Seyed Amir Kasaei, Ali Aghayari, Arash Marioriyad, Niki Sepasian, MohammadAmin Fazli, Mahdieh Soleymani Baghshah, Mohammad Hossein Rohban
https://arxiv.org/abs/2509.21227

Evaluating the Evaluators: Metrics for Compositional Text-to-Image Generation
Text-image generation has advanced rapidly, but assessing whether outputs truly capture the objects, attributes, and relations described in prompts remains a central challenge. Evaluation in this space relies heavily on automated metrics, yet these are often adopted by convention or popularity rather than validated against human judgment. Because evaluation and reported progress in the field depend directly on these metrics, it is critical to understand how well they reflect human preferences. …

🇺🇦 Auf radioeins läuft...
Metric:
🎵 Sick Muse
#NowPlaying #Metric
https://stateazure.bandcamp.com/track/metric-sick-muse-state-azure-mix
https://open.spotify.com/track/2uxLjFYEoQ7xsqfMPuwTQ9

On the Onsager-Machlup functional of the $\Phi^4$-measure
Ioannis Gasteratos, Zachary Selk
https://arxiv.org/abs/2509.20471 https://arxiv.org/pdf/2509.2047…

On the Onsager-Machlup functional of the $Φ^4$-measure
We investigate the existence of generalised densities for the $Φ^4_d$ $(d=1,2,3)$ measures, in finite volume, through the lens of Onsager-Machlup (OM) functionals. The latter are rigorously defined for measures on metric spaces as limiting ratios of small ball probabilities. In one dimension, we show that the standard OM functional of the $Φ^4_1$ measure coincides with the $Φ^4$ action as expected. In two dimensions, we show that OM functionals of the $P(Φ)_2$ measures agree with the corres…

FD-Bench: A Full-Duplex Benchmarking Pipeline Designed for Full Duplex Spoken Dialogue Systems
Yizhou Peng, Yi-Wen Chao, Dianwen Ng, Yukun Ma, Chongjia Ni, Bin Ma, Eng Siong Chng
https://arxiv.org/abs/2507.19040

FD-Bench: A Full-Duplex Benchmarking Pipeline Designed for Full Duplex Spoken Dialogue Systems
Full-duplex spoken dialogue systems (FDSDS) enable more natural human-machine interactions by allowing real-time user interruptions and backchanneling, compared to traditional SDS that rely on turn-taking. However, existing benchmarks lack metrics for FD scenes, e.g., evaluating model performance during user interruptions. In this paper, we present a comprehensive FD benchmarking pipeline utilizing LLMs, TTS, and ASR to address this gap. It assesses FDSDS's ability to handle user interruptions,…

Metrizability, connectedness, completeness and boundedness of generalized $u$-topology and $m$-topology in $C(X)$
Soumajit Dey, Sudip Kumar Acharyya, Dhananjoy Mandal
https://arxiv.org/abs/2508.14651

Metrizability, connectedness, completeness and boundedness of generalized $u$-topology and $m$-topology in $C(X)$
If $I$ is an ideal in the ring $C(X)$ of all real valued continuous functions defined over a Tychonoff space $X$, then $X$ is called $I$-$pseudocompact$ if the set $X\setminus \bigcap Z[I]$ is a bounded subset of $X$. Corresponding to $I$, the $m^I$-topology and $u^I$-topology on $C(X)$, generalizing the well-known $m$-topology and $u$-topology in $C(X)$ respectively are already there in the literature. It is proved amongst others that the $m^I$-topology is first countable if and only if the $u…

LLM Robustness Leaderboard v1 --Technical report
Pierre Peign\'e - Lefebvre, Quentin Feuillade-Montixi, Tom David, Nicolas Miailhe
https://arxiv.org/abs/2508.06296 https://

LLM Robustness Leaderboard v1 --Technical report
This technical report accompanies the LLM robustness leaderboard published by PRISM Eval for the Paris AI Action Summit. We introduce PRISM Eval Behavior Elicitation Tool (BET), an AI system performing automated red-teaming through Dynamic Adversarial Optimization that achieves 100% Attack Success Rate (ASR) against 37 of 41 state-of-the-art LLMs. Beyond binary success metrics, we propose a fine-grained robustness metric estimating the average number of attempts required to elicit harmful behav…

DeepTrust: Multi-Step Classification through Dissimilar Adversarial Representations for Robust Android Malware Detection
Daniel Pulido-Cort\'azar, Daniel Gibert, Felip Many\`a
https://arxiv.org/abs/2510.12310

DeepTrust: Multi-Step Classification through Dissimilar Adversarial Representations for Robust Android Malware Detection
Over the last decade, machine learning has been extensively applied to identify malicious Android applications. However, such approaches remain vulnerable against adversarial examples, i.e., examples that are subtly manipulated to fool a machine learning model into making incorrect predictions. This research presents DeepTrust, a novel metaheuristic that arranges flexible classifiers, like deep neural networks, into an ordered sequence where the final decision is made by a single internal model…

State Field Coverage: A Metric for Oracle Quality
Facundo Molina, Nazareno Aguirre, Alessandra Gorla
https://arxiv.org/abs/2510.03071 https://arxiv.org/pdf…

State Field Coverage: A Metric for Oracle Quality
The effectiveness of testing in uncovering software defects depends not only on the characteristics of the test inputs and how thoroughly they exercise the software, but also on the quality of the oracles used to determine whether the software behaves as expected. Therefore, assessing the quality of oracles is crucial to improve the overall effectiveness of the testing process. Existing metrics have been used for this purpose, but they either fail to provide a comprehensive basis for guiding or…

From de Sitter to Anti-de Sitter singularity regularization: theory and phenomenology
Julio Arrechea, Stefano Liberati, Hooman Neshat, Vania Vellucci
https://arxiv.org/abs/2509.13421

From de Sitter to Anti-de Sitter singularity regularization: theory and phenomenology
Recent investigations of vacuum polarization in extremely compact stars suggest that, in such regimes, the effective matter content of spacetime may acquire a vacuum-energy equation of state with negative energy density, mimicking a negative cosmological constant. Motivated by this observation, we introduce a general algorithm to modify well-known spherically symmetric regular black hole metrics by replacing their usual de Sitter cores (dSC) with Anti-de Sitter cores (AdSC). Like their dSC coun…

Vectorized 3D mesh refinement and implementation of primal hybrid FEM in MATLAB
Harish Nagula Mallesham, Sharat Gaddam, Jan Valdman, Sanjib Kumar Acharya
https://arxiv.org/abs/2509.11133

Vectorized 3D mesh refinement and implementation of primal hybrid FEM in MATLAB
In this article, we introduce a Face-to-Tetrahedron connectivity in MATLAB together with a vectorized 3D uniform mesh refinement technique. We introduce a MATLAB vectorized assembly of 3D lowest-order primal hybrid finite element matrices for a second-order elliptic problem. We introduce a parallel solver and a vectorized Schur complement solver to solve the associated linear problem. The numerical results illustrate the software's runtime performance.