Tootfinder

See the Forest and the Trees: A Synergistic Reasoning Framework for Knowledge-Based Visual Question Answering
Junjie Wang, Yunhan Tang, Yijie Wang, Zhihao Yuan, Huan Wang, Yangfan He, Bin Li
https://arxiv.org/abs/2507.17659

See the Forest and the Trees: A Synergistic Reasoning Framework for Knowledge-Based Visual Question Answering
Multimodal Large Language Models (MLLMs) have pushed the frontiers of Knowledge-Based Visual Question Answering (KBVQA), yet their reasoning is fundamentally bottlenecked by a reliance on uni-dimensional evidence. This "seeing only the trees, but not the forest" approach prevents robust, multi-faceted understanding. Inspired by the principle of seeing both the forest and trees, we propose Synergos-VQA, a novel synergistic reasoning framework. At its core, Synergos-VQA concurrently generates and…

As usual, billionaires are willing to go to great lengths to avoid having to contribute even a fraction of their wealth to the common good. These people are the root of much that is wrong with the world, and we simply need to make it both unacceptable and impossible for anyone to be a billionaire at the expense of the vast majority of people worldwide.

Inside the Jaguar Land Rover cyberattack, which forced a shutdown of most systems and is likely to cost the company hundreds of millions of pounds (The Guardian)
https://www.theguardian.com/business/2025/sep/20/jaguar-land-rover-hack-factor…

First large scale spatial and velocity patterns of local metal-rich stars in the Milky Way
Georges Kordopatis, Diane Feuillet, Christian Lehmann, Sofia Feltzing, Ivan Minchev, Vanessa Hill, Heitor Ernandes
https://arxiv.org/abs/2509.17420

First large scale spatial and velocity patterns of local metal-rich stars in the Milky Way
(Abridged) The present-day spatial and kinematic distribution of stars in the Milky Way provides key constraints on its internal dynamics and evolutionary history. We select stars that are more metal-rich than the interstellar medium (ISM) at their guiding radius (the so-called Local Metal-Rich stars, LMR) and investigate their chemo-kinematics. Until recently, existing catalogues did not contain such targets in large quantities, but one can now select many millions of them by using Gaia photom…

🤦 As Indonesia reclaims forests from palm oil, smallholders bear brunt of enforcement
https://news.mongabay.com/2025/07/as-indonesia-reclaims-forests-from-palm-oil-smallholders-bear-brunt-of-enforcement/

As Indonesia reclaims forests from palm oil, smallholders bear brunt of enforcement
JAKARTA — Indonesia has launched a sweeping and militarized crackdown on illegal forest use, reclaiming millions of hectares of land — but civil society groups warn the campaign is displacing Indigenous peoples while sparing corporate offenders and entrenching land inequality. Under a regulation issued in January, President Prabowo Subianto established a task force to crack down on […]

Implementation and Evaluation of Fast Raft for Hierarchical Consensus
Anton Melnychuk, Bryan SebaRaj
https://arxiv.org/abs/2506.17793 https://

Implementation and Evaluation of Fast Raft for Hierarchical Consensus
We present the first open-source implementation and evaluation of Fast Raft, a hierarchical consensus protocol designed for dynamic, distributed environments. Fast Raft reduces the number of message rounds needed to commit log entries compared to standard Raft by introducing a fast-track mechanism and reducing leader dependence. Our implementation uses gRPC and Kubernetes-based deployment across AWS availability zones. Experimental results demonstrate a throughput improvement and reduced commit…

WildClaims: Information Access Conversations in the Wild(Chat)
Hideaki Joko, Shakiba Amirshahi, Charles L. A. Clarke, Faegheh Hasibi
https://arxiv.org/abs/2509.17442 https://

WildClaims: Information Access Conversations in the Wild(Chat)
The rapid advancement of Large Language Models (LLMs) has transformed conversational systems into practical tools used by millions. However, the nature and necessity of information retrieval in real-world conversations remain largely unexplored, as research has focused predominantly on traditional, explicit information access conversations. The central question is: What do real-world information access conversations look like? To this end, we first conduct an observational study on the WildChat…

Interpretable Audio Editing Evaluation via Chain-of-Thought Difference-Commonality Reasoning with Multimodal LLMs
Yuhang Jia, Xu Zhang, Yang Chen, Hui Wang, Enzhi Wang, Yong Qin
https://arxiv.org/abs/2509.16975

Interpretable Audio Editing Evaluation via Chain-of-Thought Difference-Commonality Reasoning with Multimodal LLMs
Automatic mean opinion score (MOS) prediction provides a more perceptual alternative to objective metrics, offering deeper insights into the evaluated models. With the rapid progress of multimodal large language models (MLLMs), their enhanced perceptual and reasoning abilities enable more comprehensive and interpretable audio quality assessment. In this work, we tackle the challenging task of audio editing evaluation and propose the first natural language-based automated evaluation framework bu…

I’ve been mulling my own WCAG3 post, but my TL;DR on Karl’s post is essentially the same as I would have written:
“WCAG 3 is not ready for you to use as a compliance framework. Not even close.”
https://afixt.com/why-now-is-not-the-time-to-think-about-wcag-3/…

Why Now Is Not the Time to Think About WCAG 3 - AFixt
If you work in accessibility or are responsible for compliance at your organization, you’ve probably heard about WCAG 3.0. The W3C has been developing it for years, and the most recent Working Draft was released in September 2025. At first glance, it feels like a big leap forward: a standard that promises to address gaps […]

See&Trek: Training-Free Spatial Prompting for Multimodal Large Language Model
Pengteng Li, Pinhao Song, Wuyang Li, Weiyu Guo, Huizai Yao, Yijie Xu, Dugang Liu, Hui Xiong
https://arxiv.org/abs/2509.16087

See&Trek: Training-Free Spatial Prompting for Multimodal Large Language Model
We introduce SEE&TREK, the first training-free prompting framework tailored to enhance the spatial understanding of Multimodal Large Language Models (MLLMS) under vision-only constraints. While prior efforts have incorporated modalities like depth or point clouds to improve spatial reasoning, purely visualspatial understanding remains underexplored. SEE&TREK addresses this gap by focusing on two core principles: increasing visual diversity and motion reconstruction. For visual diversity, we con…