Imagen de Portada

Echoes as Anchors: Probabilistic Costs and Attention Refocusing in LLM Reasoning

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Hao, Zhuoyuan, Li, Zhuo, Li, Wu, Liu, Fangming, Zhang, Min, Li, Jing
Formato:	Preprint
Publicado:	2026
Materias:	Computation and Language
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

LaRe: Latent Refocusing for Multimodal Reasoning
por: Ma, Jizheng, et al.
Publicado: (2025)

Attention Illuminates LLM Reasoning: The Preplan-and-Anchor Rhythm Enables Fine-Grained Policy Optimization
por: Li, Yang, et al.
Publicado: (2025)

Impromptu Cybercrime Euphemism Detection
por: Li, Xiang, et al.
Publicado: (2024)

Lower Layers Matter: Alleviating Hallucination via Multi-Layer Fusion Contrastive Decoding with Truthfulness Refocused
por: Chen, Dingwei, et al.
Publicado: (2024)

Multi-objective Large Language Model Alignment with Hierarchical Experts
por: Li, Zhuo, et al.
Publicado: (2025)

Chain of Thought with Explicit Evidence Reasoning for Few-shot Relation Extraction
por: Ma, Xilai, et al.
Publicado: (2023)

ClueAnchor: Clue-Anchored Knowledge Reasoning Exploration and Optimization for Retrieval-Augmented Generation
por: Chen, Hao, et al.
Publicado: (2025)

Knowledge Editing with Dynamic Knowledge Graphs for Multi-Hop Question Answering
por: Lu, Yifan, et al.
Publicado: (2024)

Simulate, Refocus and Ensemble: An Attention-Refocusing Scheme for Domain Generalization
por: Wang, Ziyi, et al.
Publicado: (2025)

AnchorOPT: Towards Optimizing Dynamic Anchors for Adaptive Prompt Learning
por: Li, Zheng, et al.
Publicado: (2025)

ROG: Retrieval-Augmented LLM Reasoning for Complex First-Order Queries over Knowledge Graphs
por: Zhang, Ziyan, et al.
Publicado: (2026)

Attention Projection Mixing with Exogenous Anchors
por: Su, Jonathan
Publicado: (2026)

Adaptive Detoxification: Safeguarding General Capabilities of LLMs through Toxicity-Aware Knowledge Editing
por: Lu, Yifan, et al.
Publicado: (2025)

InstInfer: In-Storage Attention Offloading for Cost-Effective Long-Context LLM Inference
por: Pan, Xiurui, et al.
Publicado: (2024)

Answer-Centric or Reasoning-Driven? Uncovering the Latent Memory Anchor in LLMs
por: Wu, Yang, et al.
Publicado: (2025)

Echoes in AI: Quantifying lack of plot diversity in LLM outputs
por: Xu, Weijia, et al.
Publicado: (2024)

Thought Anchors: Which LLM Reasoning Steps Matter?
por: Bogdan, Paul C., et al.
Publicado: (2025)

NCV: A Node-Wise Consistency Verification Approach for Low-Cost Structured Error Localization in LLM Reasoning
por: Zhang, Yulong, et al.
Publicado: (2025)

Knowledge Fusion of Large Language Models Via Modular SkillPacks
por: Du, Guodong, et al.
Publicado: (2025)

Multimodal Reasoning with Multimodal Knowledge Graph
por: Lee, Junlin, et al.
Publicado: (2024)

Dual Traits in Probabilistic Reasoning of Large Language Models
por: Li, Shenxiong, et al.
Publicado: (2024)

Knowledgeable Preference Alignment for LLMs in Domain-specific Question Answering
por: Zhang, Yichi, et al.
Publicado: (2023)

Decoding the Echoes of Vision from fMRI: Memory Disentangling for Past Semantic Information
por: Xia, Runze, et al.
Publicado: (2024)

Probabilistic Soundness Guarantees in LLM Reasoning Chains
por: You, Weiqiu, et al.
Publicado: (2025)

Probabilistic Modeling of Intentions in Socially Intelligent LLM Agents
por: Xia, Feifan, et al.
Publicado: (2025)

Not All Tokens Learn Alike: Attention Entropy Reveals Heterogeneous Signals in RL Reasoning
por: Li, Gengyang, et al.
Publicado: (2026)

Power Echoes: Investigating Moderation Biases in Online Power-Asymmetric Conflicts
por: Li, Yaqiong, et al.
Publicado: (2026)

MIND Your Reasoning: A Meta-Cognitive Intuitive-Reflective Network for Dual-Reasoning in Multimodal Stance Detection
por: Wang, Bingbing, et al.
Publicado: (2025)

Do Proactive Agents Really Need an LLM to Decide When to Wake and What to Anchor?
por: Liu, Xiaoze, et al.
Publicado: (2026)

Unified Data Selection for LLM Reasoning
por: Li, Xiaoyuan, et al.
Publicado: (2026)

Instruction Anchor: Dissecting the Mechanistic Dynamics of Modality Arbitration
por: Zhang, Yu, et al.
Publicado: (2026)

Dual Reasoning: A GNN-LLM Collaborative Framework for Knowledge Graph Question Answering
por: Liu, Guangyi, et al.
Publicado: (2024)

Mediocrity is the key for LLM as a Judge Anchor Selection
por: Don-Yehiya, Shachar, et al.
Publicado: (2026)

Mathematical Reasoning Enhanced LLM for Formula Derivation: A Case Study on Fiber NLI Modellin
por: Zhang, Yao, et al.
Publicado: (2026)

Multi-Modality Expansion and Retention for LLMs through Parameter Merging and Decoupling
por: Li, Junlin, et al.
Publicado: (2025)

Beyond Two-Stage Training: Cooperative SFT and RL for LLM Reasoning
por: Chen, Liang, et al.
Publicado: (2025)

BeLLM: Backward Dependency Enhanced Large Language Model for Sentence Embeddings
por: Li, Xianming, et al.
Publicado: (2023)

Enrich-on-Graph: Query-Graph Alignment for Complex Reasoning with LLM Enriching
por: Li, Songze, et al.
Publicado: (2025)

Thinking in a Crowd: How Auxiliary Information Shapes LLM Reasoning
por: Zhao, Haodong, et al.
Publicado: (2025)

Learning Ordinal Probabilistic Reward from Preferences
por: Chen, Longze, et al.
Publicado: (2026)