Image de couverture de livre

From Growing to Looping: A Unified View of Iterative Computation in LLMs

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Kapl, Ferdinand, Angelis, Emmanouil, Maile, Kaitlin, von Oswald, Johannes, Bauer, Stefan
Format:	Preprint
Publié:	2026
Sujets:	Computation and Language Artificial Intelligence Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Do Depth-Grown Models Overcome the Curse of Depth? An In-Depth Analysis
par: Kapl, Ferdinand, et autres
Publié: (2025)

From Words to Amino Acids: Does the Curse of Depth Persist?
par: Siji, Aleena, et autres
Publié: (2026)

MesaNet: Sequence Modeling by Locally Optimal Test-Time Training
par: von Oswald, Johannes, et autres
Publié: (2025)

Iterative Deployment Improves Planning Skills in LLMs
par: Corrêa, Augusto B., et autres
Publié: (2025)

AQuA -- Combining Experts' and Non-Experts' Views To Assess Deliberation Quality in Online Discussions Using LLMs
par: Behrendt, Maike, et autres
Publié: (2024)

Evolutionary Guided Decoding: Iterative Value Refinement for LLMs
par: Liu, Zhenhua, et autres
Publié: (2025)

Memory-Efficient Looped Transformer: Decoupling Compute from Memory in Looped Language Models
par: Vendrell, Victor Conchello, et autres
Publié: (2026)

Aligning Frozen LLMs by Reinforcement Learning: An Iterative Reweight-then-Optimize Approach
par: Zhang, Xinnan, et autres
Publié: (2025)

Implicit Reward as the Bridge: A Unified View of SFT and DPO Connections
par: Wang, Bo, et autres
Publié: (2025)

Towards a Unified View of Large Language Model Post-Training
par: Lv, Xingtai, et autres
Publié: (2025)

Tracing Computation Density in LLMs
par: Kervadec, Corentin, et autres
Publié: (2026)

Stealing Creator's Workflow: A Creator-Inspired Agentic Framework with Iterative Feedback Loop for Improved Scientific Short-form Generation
par: Park, Jong Inn, et autres
Publié: (2025)

Iterative Learning of Computable Phenotypes for Treatment Resistant Hypertension using Large Language Models
par: Aldeia, Guilherme Seidyo Imai, et autres
Publié: (2025)

Large Language Model Post-Training: A Unified View of Off-Policy and On-Policy Learning
par: Zhao, Shiwan, et autres
Publié: (2026)

How do LLMs Compute Verbal Confidence
par: Kumaran, Dharshan, et autres
Publié: (2026)

A Unified Framework with Novel Metrics for Evaluating the Effectiveness of XAI Techniques in LLMs
par: Mersha, Melkamu Abay, et autres
Publié: (2025)

Unified Tool Integration for LLMs: A Protocol-Agnostic Approach to Function Calling
par: Ding, Peng, et autres
Publié: (2025)

Grow Up and Merge: Scaling Strategies for Efficient Language Adaptation
par: Glocker, Kevin, et autres
Publié: (2025)

How Post-Training Reshapes LLMs: A Mechanistic View on Knowledge, Truthfulness, Refusal, and Confidence
par: Du, Hongzhe, et autres
Publié: (2025)

When Do LLMs Reason? A Dynamical Systems View via Entropy Phase Transitions
par: Xia, Wei, et autres
Publié: (2026)

KDRL: Post-Training Reasoning LLMs via Unified Knowledge Distillation and Reinforcement Learning
par: Xu, Hongling, et autres
Publié: (2025)

View From Above: A Framework for Evaluating Distribution Shifts in Model Behavior
par: Chopra, Tanush, et autres
Publié: (2024)

LoopTool: Closing the Data-Training Loop for Robust LLM Tool Calls
par: Zhang, Kangning, et autres
Publié: (2025)

The Growing Pains of Frontier Models: When Leaderboards Stop Separating and What to Measure Next
par: Amin, Adil
Publié: (2026)

An Iterative Utility Judgment Framework Inspired by Philosophical Relevance via LLMs
par: Zhang, Hengran, et autres
Publié: (2024)

Model-in-the-Loop (MILO): Accelerating Multimodal AI Data Annotation with LLMs
par: Wang, Yifan, et autres
Publié: (2024)

HCAttention: Extreme KV Cache Compression via Heterogeneous Attention Computing for LLMs
par: Yang, Dongquan, et autres
Publié: (2025)

Reward Model Overoptimisation in Iterated RLHF
par: Wolf, Lorenz, et autres
Publié: (2025)

AQUA: Attention via QUery mAgnitudes for Memory and Compute Efficient Inference in LLMs
par: S, Santhosh G, et autres
Publié: (2025)

Towards A Unified View of Answer Calibration for Multi-Step Reasoning
par: Deng, Shumin, et autres
Publié: (2023)

From Detection to Discovery: A Closed-Loop Approach for Simultaneous and Continuous Medical Knowledge Expansion and Depression Detection on Social Media
par: Geng, Shuang, et autres
Publié: (2025)

Distinct Computations Emerge From Compositional Curricula in In-Context Learning
par: Lee, Jin Hwa, et autres
Publié: (2025)

Reasoning with Latent Thoughts: On the Power of Looped Transformers
par: Saunshi, Nikunj, et autres
Publié: (2025)

From Isolated Conversations to Hierarchical Schemas: Dynamic Tree Memory Representation for LLMs
par: Rezazadeh, Alireza, et autres
Publié: (2024)

From Associations to Activations: Comparing Behavioral and Hidden-State Semantic Geometry in LLMs
par: Schiekiera, Louis, et autres
Publié: (2026)

From Feelings to Metrics: Understanding and Formalizing How Users Vibe-Test LLMs
par: Itzhak, Itay, et autres
Publié: (2026)

LLMs are not Zero-Shot Reasoners for Biomedical Information Extraction
par: Nagar, Aishik, et autres
Publié: (2024)

Adaptive Inference-Time Compute: LLMs Can Predict if They Can Do Better, Even Mid-Generation
par: Manvi, Rohin, et autres
Publié: (2024)

Automatic Curriculum Expert Iteration for Reliable LLM Reasoning
par: Zhao, Zirui, et autres
Publié: (2024)

Iteration Head: A Mechanistic Study of Chain-of-Thought
par: Cabannes, Vivien, et autres
Publié: (2024)