Imagen de Portada

The Geometry of Self-Verification in a Task-Specific Reasoning Model

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Lee, Andrew, Sun, Lihao, Wendler, Chris, Viégas, Fernanda, Wattenberg, Martin
Formato:	Preprint
Publicado:	2025
Materias:	Artificial Intelligence Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Relational Composition in Neural Networks: A Survey and Call to Action
por: Wattenberg, Martin, et al.
Publicado: (2024)

Dialogue Action Tokens: Steering Language Models in Goal-Directed Dialogue with a Multi-Turn Planner
por: Li, Kenneth, et al.
Publicado: (2024)

When Bad Data Leads to Good Models
por: Li, Kenneth, et al.
Publicado: (2025)

Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Task
por: Li, Kenneth, et al.
Publicado: (2022)

Inference-Time Intervention: Eliciting Truthful Answers from a Language Model
por: Li, Kenneth, et al.
Publicado: (2023)

Shared Global and Local Geometry of Language Model Embeddings
por: Lee, Andrew, et al.
Publicado: (2025)

Why Can't Transformers Learn Multiplication? Reverse-Engineering Reveals Long-Range Dependency Pitfalls
por: Bai, Xiaoyan, et al.
Publicado: (2025)

LLM Reasoning as Trajectories: Step-Specific Representation Geometry and Correctness Signals
por: Sun, Lihao, et al.
Publicado: (2026)

Tensor Product Representation Probes Reveal Shared Structure Across Linear Directions
por: Lee, Andrew, et al.
Publicado: (2026)

Measuring and Controlling Instruction (In)Stability in Language Model Dialogs
por: Li, Kenneth, et al.
Publicado: (2024)

What Does it Mean for a Neural Network to Learn a "World Model"?
por: Li, Kenneth, et al.
Publicado: (2025)

Decomposing Query-Key Feature Interactions Using Contrastive Covariances
por: Lee, Andrew, et al.
Publicado: (2026)

Learning DAGs from Data with Few Root Causes
por: Misiakos, Panagiotis, et al.
Publicado: (2023)

Reasoning on the Manifold: Bidirectional Consistency for Self-Verification in Diffusion Language Models
por: Ruan, Jiaoyang, et al.
Publicado: (2026)

Does visualization help AI understand data?
por: Li, Victoria R., et al.
Publicado: (2025)

Discovering Forbidden Topics in Language Models
por: Rager, Can, et al.
Publicado: (2025)

Distill-then-Replace: Efficient Task-Specific Hybrid Attention Model Construction
por: Xia, Xiaojie, et al.
Publicado: (2026)

Reinforcement Inference: Leveraging Uncertainty for Self-Correcting Language Model Reasoning
por: Sun, Xinhai
Publicado: (2026)

SMART: Self-learning Meta-strategy Agent for Reasoning Tasks
por: Liu, Rongxing, et al.
Publicado: (2024)

ThinkSwitch: Context Distillation with LoRA and Weight Interpolation for Specific-Purpose Reasoning Tasks
por: Saini, Dhruv, et al.
Publicado: (2026)

Algorithmic Primitives and Compositional Geometry of Reasoning in Language Models
por: Lippl, Samuel, et al.
Publicado: (2025)

Think Before You Lie: How Reasoning Leads to Honesty
por: Yuan, Ann, et al.
Publicado: (2026)

One-Token Verification for Reasoning Correctness Estimation
por: Zhuang, Zhan, et al.
Publicado: (2026)

Introducing Verification Task of Set Consistency with Set-Consistency Energy Networks
por: Song, Mooho, et al.
Publicado: (2025)

ART: Adaptive Reasoning Trees for Explainable Claim Verification
por: Wadhwa, Sahil, et al.
Publicado: (2026)

Self-Verification Dilemma: Experience-Driven Suppression of Overused Checking in LLM Reasoning
por: Long, Quanyu, et al.
Publicado: (2026)

ICLR: In-Context Learning of Representations
por: Park, Core Francisco, et al.
Publicado: (2024)

Exploring Correlations of Self-Supervised Tasks for Graphs
por: Fang, Taoran, et al.
Publicado: (2024)

Large Language Models Can Self-Improve At Web Agent Tasks
por: Patel, Ajay, et al.
Publicado: (2024)

When to Trust the Cheap Check: Weak and Strong Verification for Reasoning
por: Kiyani, Shayan, et al.
Publicado: (2026)

Speculative Coreset Selection for Task-Specific Fine-tuning
por: Zhang, Xiaoyu, et al.
Publicado: (2024)

Projected Task-Specific Layers for Multi-Task Reinforcement Learning
por: Roberts, Josselin Somerville, et al.
Publicado: (2023)

A Model Can Help Itself: Reward-Free Self-Training for LLM Reasoning
por: Li, Mengqi, et al.
Publicado: (2025)

LASS-ODE: Scaling ODE Computations to Connect Foundation Models with Dynamical Physical Systems
por: Li, Haoran, et al.
Publicado: (2026)

Forward-Backward Reasoning in Large Language Models for Mathematical Verification
por: Jiang, Weisen, et al.
Publicado: (2023)

Binning as a Pretext Task: Improving Self-Supervised Learning in Tabular Domains
por: Lee, Kyungeun, et al.
Publicado: (2024)

Improving LLM Reasoning through Scaling Inference Computation with Collaborative Verification
por: Liang, Zhenwen, et al.
Publicado: (2024)

Parameterized Argumentation-based Reasoning Tasks for Benchmarking Generative Language Models
por: Steging, Cor, et al.
Publicado: (2025)

Principled Understanding of Generalization for Generative Transformer Models in Arithmetic Reasoning Tasks
por: Xu, Xingcheng, et al.
Publicado: (2024)

Information-Theoretic Limits of Safety Verification for Self-Improving Systems
por: Scrivens, Arsenios
Publicado: (2026)