paper-with-me

Papers

Reasoning in Large Language Models: A Geometric Perspective

2024-07-02 · Romain Cosentino, Sarath Shekkizhar

The advancement of large language models (LLMs) for real-world applications hinges critically on enhancing their reasoning capabilities. In this work, we explore the reasoning abilities of large language models (LLMs) through their geometrical understanding. We establish a connection between the expressive power of LLMs and the density of their self-attention graphs. Our analysis demonstrates that the density of these graphs defines the intrinsic dimension of the inputs to the MLP blocks. We demonstrate through theoretical analysis and toy examples that a higher intrinsic dimension implies a greater expressive capacity of the LLM. We further provide empirical evidence linking this geometric framework to recent advancements in methods aimed at enhancing the reasoning capabilities of LLMs.

📄 PDF Abstract BibTeX arXiv:2407.02678

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

REMA: A Unified Reasoning Manifold Framework for Interpreting Large Language Model

2025-09-26 · Bo Li, Guanzhi Deng, Ronghao Chen, Junrong Yue 외 arxiv

Understanding how Large Language Models (LLMs) perform complex reasoning and their failure mechanisms is a challenge in interpretability research. To provide a measurable geometric analysis perspective, we define the con…

Reasoning on the Manifold: Bidirectional Consistency for Self-Verification in Diffusion Language Models

2026-04-17 · Jiaoyang Ruan, Xin Gao, Yinda Chen, Hengyu Zeng 외 arxiv

While Diffusion Large Language Models (dLLMs) offer structural advantages for global planning, efficiently verifying that they arrive at correct answers via valid reasoning traces remains a critical challenge. In this wo…

SATURN: Symbolic Spatial Reasoning for Multi-Perspective Grounding

2026-06-21 · Danial Kamali, Tanawan Premsri, Shreya Rajpal, Amir Zadeh 외 arxiv

Vision-Language Models (VLMs) remain unreliable when spatial reasoning requires composing relations whose meanings depend on frames of reference. Existing neuro-symbolic methods make reasoning more explicit, but often de…

Referring ExpressionSpatial Reasoning

Predicting Camera Pose from Perspective Descriptions for Spatial Reasoning

2026-02-05 · Xuejun Zhang, Aditi Tiwari, Zhenhailong Wang, Heng Ji arxiv

Multi-image spatial reasoning remains challenging for current multimodal large language models (MLLMs). While single-view perception is inherently 2D, reasoning over multiple views requires building a coherent scene unde…

Scene UnderstandingSpatial Reasoning

Truth as a Trajectory: What Internal Representations Reveal About Large Language Model Reasoning

2026-03-01 · Hamed Damirchi, Ignacio Meza De la Jara, Ehsan Abbasnejad, Afshar Shamsi 외 arxiv

Existing explainability methods for Large Language Models (LLMs) typically treat hidden states as static points in activation space, assuming that correct and incorrect inferences can be separated using representations f…

Question Answering