Tootfinder

ContractEval: Benchmarking LLMs for Clause-Level Legal Risk Identification in Commercial Contracts
Shuang Liu, Zelong Li, Ruoyun Ma, Haiyan Zhao, Mengnan Du
https://arxiv.org/abs/2508.03080

ContractEval: Benchmarking LLMs for Clause-Level Legal Risk Identification in Commercial Contracts
The potential of large language models (LLMs) in specialized domains such as legal risk analysis remains underexplored. In response to growing interest in locally deploying open-source LLMs for legal tasks while preserving data confidentiality, this paper introduces ContractEval, the first benchmark to thoroughly evaluate whether open-source LLMs could match proprietary LLMs in identifying clause-level legal risks in commercial contracts. Using the Contract Understanding Atticus Dataset (CUAD),…

A Two-Step Learning Framework for Enhancing Sound Event Localization and Detection
Hogeon Yu
https://arxiv.org/abs/2507.22322 https://arxiv.org/pdf/2507.22…

A Two-Step Learning Framework for Enhancing Sound Event Localization and Detection
Sound Event Localization and Detection (SELD) is crucial in spatial audio processing, enabling systems to detect sound events and estimate their 3D directions. Existing SELD methods use single- or dual-branch architectures: single-branch models share SED and DoA representations, causing optimization conflicts, while dual-branch models separate tasks but limit information exchange. To address this, we propose a two-step learning framework. First, we introduce a tracwise reordering format to main…

Critical gradient optimization for quasi-isodynamic stellarators
G. T. Roberg-Clark, P. Xanthopoulos, G. G. Plunk, S. Stroteich
https://arxiv.org/abs/2506.22166

Critical gradient optimization for quasi-isodynamic stellarators
We present new and updated methods for reducing transport caused by electrostatic ion temperature gradient (ITG) driven turbulence in quasi-isodynamic (QI) configurations. We first show an updated model for the threshold (critical) gradient of localized toroidal ITG modes. It is then argued that it is desirable for ITG modes to "split" and localize in separate curvature drift wells, which is leveraged to produce a six-field-period QI configuration with a high critical gradient. We show that the…

Just saw this:
#AI can mean a lot of things these days, but lots of the popular meanings imply a bevy of harms that I definitely wouldn't feel are worth a cute fish game. In fact, these harms are so acute that even "just" playing into the AI hype becomes its own kind of harm (it's similar to blockchain in that way).
@… noticed that the authors claim the code base is 80% AI generated, which is a red flag because people with sound moral compasses wouldn't be using AI to "help" write code in the first place. The authors aren't by some miracle people who couldn't build this app without help, in case that influences your thinking about it: they have the skills to write the code themselves, although it likely would have taken longer (but also been better).
I was more interested in the fish-classification AI, and how much it might be dependent on datacenters. Thankfully, a quick glance at the code confirms they're using ONNX and running a self-trained neural network on your device. While the exponentially-increasing energy & water demands of datacenters to support billion-parameter models are a real concern, this is not that. Even a non-AI game can burn a lot of cycles on someone's phone, and I don't think there's anything to complain about energy-wise if we're just using cycles on the end user's device as long as we're not having them keep it on for hours crunching numbers like blockchain stuff does. Running whatever stuff locally while the user is playing a game is a negligible environmental concern, unlike, say, calling out to ChatGPT where you're directly feeding datacenter demand. Since they claimed to have trained the network themselves, and since it's actually totally reasonable to make your own dataset for this and get good-enough-for-a-silly-game results with just a few hundred examples, I don't have any ethical objections to the data sourcing or training processes either. Hooray! This is finally an example of "ethical use of neutral networks" that I can hold up as an example of what people should be doing instead of the BS they are doing.
But wait... Remember what I said about feeding the AI hype being its own form of harm? Yeah, between using AI tools for coding and calling their classifier "AI" in a way that makes it seem like the same kind of thing as ChatGPT et al., they're leaning into the hype rather than helping restrain it. And that means they're causing harm. Big AI companies can point to them and say "look AI enables cute things you like" when AI didn't actually enable it. So I'm feeling meh about this cute game and won't be sharing it aside from this post. If you love the cute fish, you don't really have to feel bad for playing with it, but I'd feel bad for advertising it without a disclaimer.

Revisiting the configurations of hydrogen impurities in SrTiO3: Insights from first-principles local vibration mode calculations
Cai Zenghua, Ma Chunlan
https://arxiv.org/abs/2507.05752

Revisiting the configurations of hydrogen impurities in SrTiO3: Insights from first-principles local vibration mode calculations
The specific configurations of hydrogen impurities in SrTiO3 (STO) are still ambiguous. In this study, we systematically investigate the configurations and vibrational properties of hy-drogen impurities in cubic STO using first-principles local vibration mode calculations. Em-ploying the appropriate hybrid exchange correlation functional with the fraction of exact ex-change setting to 0.2, we revisit the interstitial hydrogen (Hi), Hi complexes (2Hi), and various intrinsic cation vacancy comple…

Some global properties of umbilic points of Willmore immersions in the $3$-sphere
Nicolas Marque, Dorian Martino
https://arxiv.org/abs/2506.10720 https://

Some global properties of umbilic points of Willmore immersions in the $3$-sphere
We study the umbilic points of Willmore surfaces in codimension 1 from the viewpoint of the conformal Gauss map. We first study the local behaviour of the conformal Gauss map near umbilic curves and prove that they are geodesics up to a conformal transformation if and only if the Willmore immersion is, up to a conformal transformation, the gluing of minimal surfaces in the 3-dimensional hyperbolic space. Then, we prove a Gauss--Bonnet formula for the conformal Gauss map of Willmore surfaces whi…

Decadal evolution of a repeating fast radio burst source
P. Wang, J. S. Zhang, Y. P. Yang, D. K. Zhou, Y. K. Zhang, Y. Feng, Z. Y. Zhao, J. H. Fang, D. Li, W. W. Zhu, B. Zhang, F. Y. Wang, Y. F. Huang, R. Luo, J. L. Han, K. J. Lee, C. W. Tsai, Z. G. Dai, H. Gao, X. P. Zheng, J. H. Cao, X. L. Chen, E. Gugercinoglu, J. C. Jiang, W. C. Jing, Y. Li, J. Li, W. J. Lu, J. W. Luo, F. Lyu, C. C. Miao, C. H. Niu, J. R. Niu, Y. Qu, W. Y. Wang, Y. D. Wang, Y. B. Wang, C. J. Wang, Q. Wu, Y. S. Wu, …

Decadal evolution of a repeating fast radio burst source
The origin of fast radio bursts (FRBs), the brightest cosmic radio explosions, is still unknown. Bearing critical clues to FRBs' origin, the long-term evolution of FRBs has yet to be confirmed, since the field is still young and most FRBs were seen only once. Here we report clear evidence of decadal evolution of FRB~20121102A, the first precisely localized repeater. In conjunction with archival data, our FAST and GBT monitoring campaign since 2020 reveals a significant 7% decline of local dispe…

Physics-guided impact localisation and force estimation in composite plates with uncertainty quantification
Dong Xiao, Zahra Sharif-Khodaei, M. H. Aliabadi
https://arxiv.org/abs/2507.13376

Physics-guided impact localisation and force estimation in composite plates with uncertainty quantification
Physics-guided approaches offer a promising path toward accurate and generalisable impact identification in composite structures, especially when experimental data are sparse. This paper presents a hybrid framework for impact localisation and force estimation in composite plates, combining a data-driven implementation of First-Order Shear Deformation Theory (FSDT) with machine learning and uncertainty quantification. The structural configuration and material properties are inferred from dispers…

Neel order, spin-spiral, and spin liquid ground state in frustrated three dimensional system CaMn2P2: A DFT U and spin dynamics study
Bidyut Mallick, Sk. Soyeb Ali, S. K. Panda
https://arxiv.org/abs/2506.10254

Neel order, spin-spiral, and spin liquid ground state in frustrated three dimensional system CaMn2P2: A DFT+U and spin dynamics study
We investigate the magnetic ground state and phase transitions in the frustrated three-dimensional system CaMn2P2 using first-principles calculations combined with spin-dynamics simulations. Our DFT+U calculations reveal that CaMn2P2 exhibits an indirect gap semiconducting ground state with a localized Mn2+ electronic configuration and negligible spin-orbit coupling effects. The computed exchange interactions show that the magnetic behavior is well described by a isotropic Heisenberg Hamiltonia…