Cover Image

On the Failure of Latent State Persistence in Large Language Models

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Huang, Jen-tse, Sun, Kaiser, Wang, Wenxuan, Dredze, Mark
Format:	Preprint
Published:	2025
Subjects:	Computation and Language Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Amuro and Char: Analyzing the Relationship between Pre-Training and Fine-Tuning of Large Language Models
by: Sun, Kaiser, et al.
Published: (2024)

All Languages Matter: On the Multilingual Safety of Large Language Models
by: Wang, Wenxuan, et al.
Published: (2023)

Not All Countries Celebrate Thanksgiving: On the Cultural Dominance in Large Language Models
by: Wang, Wenxuan, et al.
Published: (2023)

Task Matters: Knowledge Requirements Shape LLM Responses to Context-Memory Conflict
by: Sun, Kaiser, et al.
Published: (2025)

RAG LLMs are Not Safer: A Safety Analysis of Retrieval-Augmented Generation for Large Language Models
by: An, Bang, et al.
Published: (2025)

Probing Multimodal Large Language Models on Cognitive Biases in Chinese Short-Video Misinformation
by: Huang, Jen-tse, et al.
Published: (2026)

Can one size fit all?: Measuring Failure in Multi-Document Summarization Domain Transfer
by: DeLucia, Alexandra, et al.
Published: (2025)

Towards Evaluating Proactive Risk Awareness of Multimodal Language Models
by: Yuan, Youliang, et al.
Published: (2025)

On the Shortcut Learning in Multilingual Neural Machine Translation
by: Wang, Wenxuan, et al.
Published: (2024)

LogicAsker: Evaluating and Improving the Logical Reasoning Ability of Large Language Models
by: Wan, Yuxuan, et al.
Published: (2024)

Identifying the Achilles' Heel: An Iterative Method for Dynamically Uncovering Factual Errors in Large Language Models
by: Wang, Wenxuan, et al.
Published: (2024)

Refuse Whenever You Feel Unsafe: Improving Safety in LLMs via Decoupled Refusal Training
by: Yuan, Youliang, et al.
Published: (2024)

Evaluating Implicit Biases in LLM Reasoning through Logic Grid Puzzles
by: Jahara, Fatima, et al.
Published: (2025)

Testing and Evaluation of Large Language Models: Correctness, Non-Toxicity, and Fairness
by: Wang, Wenxuan
Published: (2024)

Chain-of-Jailbreak Attack for Image Generation Models via Editing Step by Step
by: Wang, Wenxuan, et al.
Published: (2024)

Can't See the Forest for the Trees: Benchmarking Multimodal Safety Awareness for Multimodal LLMs
by: Wang, Wenxuan, et al.
Published: (2025)

How Far Are We on the Decision-Making of LLMs? Evaluating LLMs' Gaming Ability in Multi-Agent Environments
by: Huang, Jen-tse, et al.
Published: (2024)

Artificial Intolerance: Stigmatizing Language in Clinical Documentation Skews Large Language Model Decision-Making
by: Huang, Jen-tse, et al.
Published: (2026)

New Job, New Gender? Measuring the Social Bias in Image Generation Models
by: Wang, Wenxuan, et al.
Published: (2024)

ComboBench: Can LLMs Manipulate Physical Devices to Play Virtual Reality Games?
by: Li, Shuqing, et al.
Published: (2025)

No Reliable Evidence of Self-Reported Sentience in Small Large Language Models
by: Kaiser, Caspar, et al.
Published: (2026)

Compressing Sequences in the Latent Embedding Space: $K$-Token Merging for Large Language Models
by: Xu, Zihao, et al.
Published: (2026)

Dissecting Failure Dynamics in Large Language Model Reasoning
by: Zhu, Wei, et al.
Published: (2026)

Legal Documents Drafting with Fine-Tuned Pre-Trained Large Language Model
by: Lin, Chun-Hsien, et al.
Published: (2024)

Pruning General Large Language Models into Customized Expert Models
by: Zhao, Yirao, et al.
Published: (2025)

Learning to Ask: When LLM Agents Meet Unclear Instruction
by: Wang, Wenxuan, et al.
Published: (2024)

How do Large Language Models Handle Multilingualism?
by: Zhao, Yiran, et al.
Published: (2024)

Large Language Models Are Latent Variable Models: Explaining and Finding Good Demonstrations for In-Context Learning
by: Wang, Xinyi, et al.
Published: (2023)

Efficient Post-Training Refinement of Latent Reasoning in Large Language Models
by: Wang, Xinyuan, et al.
Published: (2025)

MixCE: Training Autoregressive Language Models by Mixing Forward and Reverse Cross-Entropies
by: Zhang, Shiyue, et al.
Published: (2023)

TrustMH-Bench: A Comprehensive Benchmark for Evaluating the Trustworthiness of Large Language Models in Mental Health
by: Xiong, Zixin, et al.
Published: (2026)

Control Illusion: The Failure of Instruction Hierarchies in Large Language Models
by: Geng, Yilin, et al.
Published: (2025)

Offset Unlearning for Large Language Models
by: Huang, James Y., et al.
Published: (2024)

Large Language Model Reasoning Failures
by: Song, Peiyang, et al.
Published: (2026)

SDE-SQL: Enhancing Text-to-SQL Generation in Large Language Models via Self-Driven Exploration with SQL Probes
by: Xie, Wenxuan, et al.
Published: (2025)

AdaMergeX: Cross-Lingual Transfer with Large Language Models via Adaptive Adapter Merging
by: Zhao, Yiran, et al.
Published: (2024)

Unlocking the Working Memory of Large Language Models for Latent Reasoning
by: Aichberger, Lukas, et al.
Published: (2026)

On Effects of Steering Latent Representation for Large Language Model Unlearning
by: Huu-Tien, Dang, et al.
Published: (2024)

Meanings and Feelings of Large Language Models: Observability of Latent States in Generative AI
by: Liu, Tian Yu, et al.
Published: (2024)

Character as a Latent Variable in Large Language Models: A Mechanistic Account of Emergent Misalignment and Conditional Safety Failures
by: Su, Yanghao, et al.
Published: (2026)