Tootfinder

The Role of Exploration Modules in Small Language Models for Knowledge Graph Question Answering
Yi-Jie Cheng, Oscar Chew, Yun-Nung Chen
https://arxiv.org/abs/2509.07399 https://…

The Role of Exploration Modules in Small Language Models for Knowledge Graph Question Answering
Integrating knowledge graphs (KGs) into the reasoning processes of large language models (LLMs) has emerged as a promising approach to mitigate hallucination. However, existing work in this area often relies on proprietary or extremely large models, limiting accessibility and scalability. In this study, we investigate the capabilities of existing integration methods for small language models (SLMs) in KG-based question answering and observe that their performance is often constrained by their l…

Toric geometry of ReLU neural networks
Yaoying Fu
https://arxiv.org/abs/2509.05894 https://arxiv.org/pdf/2509.05894…

Toric geometry of ReLU neural networks
Given a continuous finitely piecewise linear function $f:\mathbb{R}^{n_0} \to \mathbb{R}$ and a fixed architecture $(n_0,\ldots,n_k;1)$ of feedforward ReLU neural networks, the exact function realization problem is to determine when some network with the given architecture realizes $f$. To develop a systematic way to answer these questions, we establish a connection between toric geometry and ReLU neural networks. This approach enables us to utilize numerous structures and tools from algebraic …

Reykjavík Grapevine provides answers to questions you were too afraid to ask
https://grapevine.is/mag/2025/07/07/help-call-charlie-campsites-expensive-vegetables-beautiful-but-harmless-eruption-more/…

From Iceland — Help? Call Charlie! Campsites, Expensive Vegetables, Beautiful But Harmless Eruption & More
Unsolicited advice from an unreliable advisor Help is a four-letter verb meaning “to make it easier or possible for someone...

In an internal Meta essay, an outgoing AI researcher says a "culture of fear", with performance reviews and layoffs, is spreading like "a metastatic cancer" (Kalley Huang/The Information)
https://www.theinformation.com/articles/me…

Meta AI Researcher Warns of ‘Metastatic Cancer’ Afflicting Company Culture
Over the past month, Meta Platforms has overhauled its stumbling efforts in artificial intelligence by making a head-spinning series of hires of leaders and researchers from outside the company. But a key question has lingered: Why has Meta struggled so badly in AI that it’s injecting so much ...

A proposal for the role of higher spin fields in Conformal Cyclic Cosmology
Anders K. H. Bengtsson
https://arxiv.org/abs/2507.06634 https://

A proposal for the role of higher spin fields in Conformal Cyclic Cosmology
This paper proposes a research program into how the special properties of the cubic chiral higher spin gravity theory may play a role in conformal cyclic cosmology and black hole physics, in particular for the black hole information paradox. The paper is qualitative, raising a number of questions and connections, that may be interesting to pursue in more detail. The proposal is based on a new -- at least not so commonly expressed -- view on the presumptive role of higher spin gauge fields in fu…

Quadratic units and cubic fields
Florian Breuer, James Punch
https://arxiv.org/abs/2507.06579 https://arxiv.org/pdf/2507.06579…

Quadratic units and cubic fields
We investigate Eisenstein discriminants, which are squarefree integers $d \equiv 5 \pmod{8}$ such that the fundamental unit $\varepsilon_d$ of the real quadratic field $K=\mathbb{Q}(\sqrt{d})$ satisfies $\varepsilon_d \equiv 1 \pmod{2\mathcal{O}_K}$. These discriminants are related to a classical question of Eisenstein and have connections to the class groups of orders in quadratic fields as well as to real cubic fields. We present numerical computations of Eisenstein discriminants up to $10^{1…

Unveiling the Listener Structure Underlying K-pop's Global Success: A Large-Scale Listening Data Analysis
Ryota Nakamura, Keita Nishimoto, Ichiro Sakata, Kimitaka Asatani
https://arxiv.org/abs/2509.06606

Unveiling the Listener Structure Underlying K-pop's Global Success: A Large-Scale Listening Data Analysis
From the mid-2000s to the 2010s, K-pop moved beyond its status as a regionally popular genre in Asia and established itself as a global music genre with enthusiastic fans around the world. However, little is known about how the vast number of music listeners across the globe have listened to and perceived K-pop. This study addresses this question by analyzing a large-scale listening dataset from Last.fm. An analysis of the distribution of play counts reveals that K-pop experienced a significant…

Avoiding Knowledge Edit Skipping in Multi-hop Question Answering with Guided Decomposition
Yi Liu, Xiangrong Zhu, Xiangyu Liu, Wei Wei, Wei Hu
https://arxiv.org/abs/2509.07555 h…

Avoiding Knowledge Edit Skipping in Multi-hop Question Answering with Guided Decomposition
In a rapidly evolving world where information updates swiftly, knowledge in large language models (LLMs) becomes outdated quickly. Retraining LLMs is not a cost-effective option, making knowledge editing (KE) without modifying parameters particularly necessary. We find that although existing retrieval-augmented generation (RAG)-based KE methods excel at editing simple knowledge, they struggle with KE in multi-hop question answering due to the issue of "edit skipping", which refers to skipping t…

KERAG: Knowledge-Enhanced Retrieval-Augmented Generation for Advanced Question Answering
Yushi Sun, Kai Sun, Yifan Ethan Xu, Xiao Yang, Xin Luna Dong, Nan Tang, Lei Chen
https://arxiv.org/abs/2509.04716

KERAG: Knowledge-Enhanced Retrieval-Augmented Generation for Advanced Question Answering
Retrieval-Augmented Generation (RAG) mitigates hallucination in Large Language Models (LLMs) by incorporating external data, with Knowledge Graphs (KGs) offering crucial information for question answering. Traditional Knowledge Graph Question Answering (KGQA) methods rely on semantic parsing, which typically retrieves knowledge strictly necessary for answer generation, thus often suffer from low coverage due to rigid schema requirements and semantic ambiguity. We present KERAG, a novel KG-based…

Enhancing Food-Domain Question Answering with a Multimodal Knowledge Graph: Hybrid QA Generation and Diversity Analysis
Srihari K B, Pushpak Bhattacharyya
https://arxiv.org/abs/2507.06571

Enhancing Food-Domain Question Answering with a Multimodal Knowledge Graph: Hybrid QA Generation and Diversity Analysis
We propose a unified food-domain QA framework that combines a large-scale multimodal knowledge graph (MMKG) with generative AI. Our MMKG links 13,000 recipes, 3,000 ingredients, 140,000 relations, and 14,000 images. We generate 40,000 QA pairs using 40 templates and LLaVA/DeepSeek augmentation. Joint fine-tuning of Meta LLaMA 3.1-8B and Stable Diffusion 3.5-Large improves BERTScore by 16.2\%, reduces FID by 37.8\%, and boosts CLIP alignment by 31.1\%. Diagnostic analyses-CLIP-based mismatch det…