Tootfinder

Learning the Wrong Lessons: Syntactic-Domain Spurious Correlations in Language Models
Chantal Shaib, Vinith M. Suriyakumar, Levent Sagun, Byron C. Wallace, Marzyeh Ghassemi
https://arxiv.org/abs/2509.21155

Learning the Wrong Lessons: Syntactic-Domain Spurious Correlations in Language Models
For an LLM to correctly respond to an instruction it must understand both the semantics and the domain (i.e., subject area) of a given task-instruction pair. However, syntax can also convey implicit information Recent work shows that syntactic templates--frequent sequences of Part-of-Speech (PoS) tags--are prevalent in training data and often appear in model outputs. In this work we characterize syntactic templates, domain, and semantics in task-instruction pairs. We identify cases of spurious …

Surya - Foundation Model for Heliophysics / SuryaBench - Benchmark Dataset for Advancing Machine Learning in Heliophysics and #SpaceWeather Prediction: https://arxiv.org/abs/2508.14112 / https://arxiv.org/abs/2508.14107 -> Here comes a foundation model for the Sun: https://research.ibm.com/blog/surya-heliophysics-ai-model-sun?previewToken=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpZCI6MTEsImlhdCI6MTc1NTYyMTUwOSwiZXhwIjoxNzU1ODgwNzA5LCJzdWIiOiI0NDY1In0.pv2-edPifhET2uDPQB4hsp4HsBzhZycvMcFlZpjFQKA -> NASA and IBM Unveil AI That Helps Scientists Forecast Solar Storms: https://gizmodo.com/nasa-and-ibm-unveil-ai-that-helps-scientists-forecast-solar-storms-2000645123 - experts recently identified major gaps in their ability to forecast space weather; Surya aims to fill them.

First-Order LTLf Synthesis with Lookback (Extended Version)
Sarah Winkler
https://arxiv.org/abs/2508.18149 https://arxiv.org/pdf/2508.18149

First-Order LTLf Synthesis with Lookback (Extended Version)
Reactive synthesis addresses the problem of generating a controller for a temporal specification in an adversarial environment; it was typically studied for LTL. Driven by applications ranging from AI to business process management, LTL modulo first order-theories over finite traces (LTLfMT) has recently gained traction, where propositional variables in properties are replaced by first-order constraints. Though reactive synthesis for LTLf with some first-order features has been addressed, exist…

2025 NFL win total projections for all 32 teams: Experts react to our model https://www.nytimes.com/athletic/6572327/2025/08/25/nfl-2025-win-totals-projections-analysis/

🇺🇦 #NowPlaying on KEXP's #SundaySoul
Eko:
🎵 Kilimandjaro My Home
#Eko
https://cometrecords.bandcamp.com/track/eko-kilimandjaro-my-home
https://open.spotify.com/track/2MDlIo4i44Y88jCxUU1fl4

Negative Shanshui: Real-time Interactive Ink Painting Synthesis
Aven-Le Zhou
https://arxiv.org/abs/2508.16612 https://arxiv.org/pdf/2508.16612

Negative Shanshui: Real-time Interactive Ink Painting Synthesis
This paper presents Negative Shanshui, a real-time interactive AI synthesis approach that reinterprets classical Chinese landscape ink painting, i.e., shanshui, to engage with ecological crises in the Anthropocene. Negative Shanshui optimizes a fine-tuned Stable Diffusion model for real-time inferences and integrates it with gaze-driven inpainting, frame interpolation; it enables dynamic morphing animations in response to the viewer's gaze and presents as an interactive virtual reality (VR) exp…

Decipher-MR: A Vision-Language Foundation Model for 3D MRI Representations
Zhijian Yang, Noel DSouza, Istvan Megyeri, Xiaojian Xu, Amin Honarmandi Shandiz, Farzin Haddadpour, Krisztian Koos, Laszlo Rusko, Emanuele Valeriano, Bharadwaj Swaninathan, Lei Wu, Parminder Bhatia, Taha Kass-Hout, Erhan Bas
https://arxiv.org/abs/2509.21249

Decipher-MR: A Vision-Language Foundation Model for 3D MRI Representations
Magnetic Resonance Imaging (MRI) is a critical medical imaging modality in clinical diagnosis and research, yet its complexity and heterogeneity pose challenges for automated analysis, particularly in scalable and generalizable machine learning applications. While foundation models have revolutionized natural language and vision tasks, their application to MRI remains limited due to data scarcity and narrow anatomical focus. In this work, we present Decipher-MR, a 3D MRI-specific vision-languag…

Analysis of Circuit-based Per-Panel Diode Model of Photovoltaic Array
Peng Sang, Santhosh Balasubramanian, Amritanshu Pandey
https://arxiv.org/abs/2508.17374 https://

Analysis of Circuit-based Per-Panel Diode Model of Photovoltaic Array
Solar photovoltaic systems are increasing in size and number on the grid. In regions with high penetration, such as California, PV systems serve multiple functions, including peak shaving and demand response. Therefore, the criticality of PV systems to grid operations calls for accurate models. The current practice is to represent the PV array, composed of multiple PV panels, with an aggregated single-diode model (SDM). The highly abstract model has a limited ability to capture real-world behav…

2025 NFL win total projections for all 32 teams: Experts react to our model https://www.nytimes.com/athletic/6572327/2025/08/25/nfl-2025-win-totals-projections-analysis/

Synthesizing Artifact Dataset for Pixel-level Detection
Dennis Menn, Feng Liang, Diana Marculescu
https://arxiv.org/abs/2509.19589 https://arxiv.org/pdf/25…

Synthesizing Artifact Dataset for Pixel-level Detection
Artifact detectors have been shown to enhance the performance of image-generative models by serving as reward models during fine-tuning. These detectors enable the generative model to improve overall output fidelity and aesthetics. However, training the artifact detector requires expensive pixel-level human annotations that specify the artifact regions. The lack of annotated data limits the performance of the artifact detector. A naive pseudo-labeling approach-training a weak detector and using…