Cover Image

Controlling the Risk of Corrupted Contexts for Language Models via Early-Exiting

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Wynn, Andrea, Jazbec, Metod, Peris, Charith, Khaziev, Rinat, Liu, Anqi, Khashabi, Daniel, Nalisnick, Eric
Format:	Preprint
Published:	2025
Subjects:	Artificial Intelligence Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Fast yet Safe: Early-Exiting with Risk Control
by: Jazbec, Metod, et al.
Published: (2024)

Monitoring Risks in Test-Time Adaptation
by: Schirmer, Mona, et al.
Published: (2025)

Early-Exit Neural Networks with Nested Prediction Sets
by: Jazbec, Metod, et al.
Published: (2023)

Generative Uncertainty in Diffusion Models
by: Jazbec, Metod, et al.
Published: (2025)

Efficient and Uncertainty-Aware Diffusion Framework for Offline-to-Online Reinforcement Learning
by: Bui, Ha Manh, et al.
Published: (2026)

Regularization Through Reasoning: Systematic Improvements in Language Model Classification via Explanation-Enhanced Fine-Tuning
by: Shah, Vivswan, et al.
Published: (2025)

Constrained Entropic Unlearning: A Primal-Dual Framework for Large Language Models
by: Entesari, Taha, et al.
Published: (2025)

Dynamic Vocabulary Pruning in Early-Exit LLMs
by: Vincenti, Jort, et al.
Published: (2024)

Conformal Thinking: Risk Control for Reasoning on a Compute Budget
by: Wang, Xi, et al.
Published: (2026)

Steered LLM Activations are Non-Surjective
by: Mishra, Aayush, et al.
Published: (2026)

IA2: Alignment with ICL Activations Improves Supervised Fine-Tuning
by: Mishra, Aayush, et al.
Published: (2025)

Genomic Next-Token Predictors are In-Context Learners
by: Breslow, Nathan, et al.
Published: (2025)

Performance Control in Early Exiting to Deploy Large Models at the Same Cost of Smaller Ones
by: Mofakhami, Mehrnaz, et al.
Published: (2024)

Nav-EE: Navigation-Guided Early Exiting for Efficient Vision-Language Models in Autonomous Driving
by: Hu, Haibo, et al.
Published: (2025)

SpecEE: Accelerating Large Language Model Inference with Speculative Early Exiting
by: Xu, Jiaming, et al.
Published: (2025)

GEAR: Augmenting Language Models with Generalizable and Efficient Tool Resolution
by: Lu, Yining, et al.
Published: (2023)

AD-EE: Early Exiting for Fast and Reliable Vision-Language Models in Autonomous Driving
by: Huang, Lianming, et al.
Published: (2025)

ATOD: An Evaluation Framework and Benchmark for Agentic Task-Oriented Dialogue Systems
by: Zhang, Yifei, et al.
Published: (2026)

Are vision language models robust to uncertain inputs?
by: Wang, Xi, et al.
Published: (2025)

The Myth of Expert Specialization in MoEs: Why Routing Reflects Geometry, Not Necessarily Domain Expertise
by: Wang, Xi, et al.
Published: (2026)

Improving Prediction Certainty Estimation for Reliable Early Exiting via Null Space Projection
by: He, Jianing, et al.
Published: (2025)

K-Edit: Language Model Editing with Contextual Knowledge Awareness
by: Markowitz, Elan, et al.
Published: (2025)

Do pretrained Transformers Learn In-Context by Gradient Descent?
by: Shen, Lingfeng, et al.
Published: (2023)

ConsistentEE: A Consistent and Hardness-Guided Early Exiting Method for Accelerating Language Models Inference
by: Zeng, Ziqian, et al.
Published: (2023)

COSEE: Consistency-Oriented Signal-Based Early Exiting via Calibrated Sample Weighting Mechanism
by: He, Jianing, et al.
Published: (2024)

Sustainable Edge Intelligence Through Energy-Aware Early Exiting
by: Bullo, Marcello, et al.
Published: (2023)

Tree-of-Traversals: A Zero-Shot Reasoning Algorithm for Augmenting Black-box Language Models with Knowledge Graphs
by: Markowitz, Elan, et al.
Published: (2024)

A Tale of Two Temperatures: Simple, Efficient, and Diverse Sampling from Diffusion Language Models
by: Olausson, Theo X., et al.
Published: (2026)

Temporal Test-Time Adaptation with State-Space Models
by: Schirmer, Mona, et al.
Published: (2024)

Dynamic Mixture-of-Experts for Visual Autoregressive Model
by: Vincenti, Jort, et al.
Published: (2025)

Principled Context Engineering for RAG: Statistical Guarantees via Conformal Prediction
by: Chakraborty, Debashish, et al.
Published: (2025)

Early Exiting Predictive Coding Neural Networks for Edge AI
by: Zniber, Alaa, et al.
Published: (2023)

Large Language Model Unlearning via Embedding-Corrupted Prompts
by: Liu, Chris Yuhao, et al.
Published: (2024)

Emergent Strategic Reasoning Risks in AI: A Taxonomy-Driven Evaluation Framework
by: Kumarage, Tharindu, et al.
Published: (2026)

Geometry over Density: Few-Shot Cross-Domain OOD Detection
by: Li, Shawn, et al.
Published: (2026)

The Language Barrier: Dissecting Safety Challenges of LLMs in Multilingual Contexts
by: Shen, Lingfeng, et al.
Published: (2024)

Defenses Against Prompt Attacks Learn Surface Heuristics
by: Li, Shawn, et al.
Published: (2026)

Generative Modeling from Black-box Corruptions via Self-Consistent Stochastic Interpolants
by: Modi, Chirag, et al.
Published: (2025)

"According to ...": Prompting Language Models Improves Quoting from Pre-Training Data
by: Weller, Orion, et al.
Published: (2023)

Controllable Safety Alignment: Inference-Time Adaptation to Diverse Safety Requirements
by: Zhang, Jingyu, et al.
Published: (2024)