Copertina

Rainbow Padding: Mitigating Early Termination in Instruction-Tuned Diffusion LLMs

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Kim, Bumjun, Jeon, Dongjae, Kim, Dueun, Jeung, Wonje, No, Albert
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Artificial Intelligence
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Understanding and Mitigating Memorization in Generative Models via Sharpness of Probability Landscapes
di: Jeon, Dongjae, et al.
Pubblicazione: (2024)

An Information Theoretic Evaluation Metric For Strong Unlearning
di: Jeon, Dongjae, et al.
Pubblicazione: (2024)

Few-Shot Truly Benign DPO Attack for Jailbreaking LLMs
di: Yoon, Sangyeon, et al.
Pubblicazione: (2026)

VLMs Trace Without Tracking: Diagnosing Failures in Visual Path Following
di: Hong, Hyesoo, et al.
Pubblicazione: (2026)

The Confidence Shortcut: A Reasoning Failure Mode of Masked Diffusion Models
di: Kim, Dueun, et al.
Pubblicazione: (2026)

A2D: Any-Order, Any-Step Safety Alignment for Diffusion Language Models
di: Jeung, Wonje, et al.
Pubblicazione: (2025)

Multi-Level Knowledge Distillation and Dynamic Self-Supervised Learning for Continual Learning
di: Kim, Taeheon, et al.
Pubblicazione: (2025)

DAPD: Dependency-Aware Parallel Decoding via Attention for Diffusion LLMs
di: Kim, Bumjun, et al.
Pubblicazione: (2026)

SAFEPATH: Preventing Harmful Reasoning in Chain-of-Thought via Early Alignment
di: Jeung, Wonje, et al.
Pubblicazione: (2025)

Memorization In Stable Diffusion Is Unexpectedly Driven by CLIP Embeddings
di: Kim, Bumjun, et al.
Pubblicazione: (2026)

ReALFRED: An Embodied Instruction Following Benchmark in Photo-Realistic Environments
di: Kim, Taewoong, et al.
Pubblicazione: (2024)

Preserve-Then-Quantize: Balancing Rank Budgets for Quantization Error Reconstruction in LLMs
di: Cho, Yoonjun, et al.
Pubblicazione: (2026)

Rethinking Benign Relearning: Syntax as the Hidden Driver of Unlearning Failures
di: Yoon, Sangyeon, et al.
Pubblicazione: (2026)

BenchPreS: A Benchmark for Context-Aware Personalized Preference Selectivity of Persistent-Memory LLMs
di: Yoon, Sangyeon, et al.
Pubblicazione: (2026)

Large Language Models Still Exhibit Bias in Long Text
di: Jeung, Wonje, et al.
Pubblicazione: (2024)

SEPS: A Separability Measure for Robust Unlearning in LLMs
di: Jeung, Wonje, et al.
Pubblicazione: (2025)

Assigning Distinct Roles to Quantized and Low-Rank Matrices Toward Optimal Weight Decomposition
di: Cho, Yoonjun, et al.
Pubblicazione: (2025)

A Theoretical Analysis of Why Masked Diffusion Models Mitigate the Reversal Curse
di: Jeon, Moongyu, et al.
Pubblicazione: (2026)

NeSyPr: Neurosymbolic Proceduralization For Efficient Embodied Reasoning
di: Choi, Wonje, et al.
Pubblicazione: (2025)

EDIT: Early Diffusion Inference Termination for dLLMs Based on Dynamics of Training Gradients
di: Hsieh, He-Yen, et al.
Pubblicazione: (2025)

Cross-Domain Demo-to-Code via Neurosymbolic Counterfactual Reasoning
di: Kim, Jooyoung, et al.
Pubblicazione: (2026)

R-TOFU: Unlearning in Large Reasoning Models
di: Yoon, Sangyeon, et al.
Pubblicazione: (2025)

Adversarial Sample-Based Approach for Tighter Privacy Auditing in Final Model-Only Scenarios
di: Yoon, Sangyeon, et al.
Pubblicazione: (2024)

TTA-DAME: Test-Time Adaptation with Domain Augmentation and Model Ensemble for Dynamic Driving Conditions
di: Jeon, Dongjae, et al.
Pubblicazione: (2025)

Embodied CoT Distillation From LLM To Off-the-shelf Agents
di: Choi, Wonje, et al.
Pubblicazione: (2024)

Learning Equi-angular Representations for Online Continual Learning
di: Seo, Minhyuk, et al.
Pubblicazione: (2024)

MOGAM: A Multimodal Object-oriented Graph Attention Model for Depression Detection
di: Cha, Junyeop, et al.
Pubblicazione: (2024)

Diffusion Instruction Tuning
di: Jin, Chen, et al.
Pubblicazione: (2025)

Text2Chart31: Instruction Tuning for Chart Generation with Automatic Feedback
di: Zadeh, Fatemeh Pesaran, et al.
Pubblicazione: (2024)

Efficient Policy Adaptation with Contrastive Prompt Ensemble for Embodied Agents
di: Choi, Wonje, et al.
Pubblicazione: (2024)

Hallucination by Code Generation LLMs: Taxonomy, Benchmarks, Mitigation, and Challenges
di: Lee, Yunseo, et al.
Pubblicazione: (2025)

SynAD: Enhancing Real-World End-to-End Autonomous Driving Models through Synthetic Data Integration
di: Kim, Jongsuk, et al.
Pubblicazione: (2025)

A Systematic Evaluation of Parameter-Efficient Fine-Tuning Methods for the Security of Code LLMs
di: Lee, Kiho, et al.
Pubblicazione: (2025)

EXAONE 3.0 7.8B Instruction Tuned Language Model
di: An, Soyoung, et al.
Pubblicazione: (2024)

Unifying Block-wise PTQ and Distillation-based QAT for Progressive Quantization toward 2-bit Instruction-Tuned LLMs
di: Lee, Jung Hyun, et al.
Pubblicazione: (2025)

Two-Stage Grid Optimization for Group-wise Quantization of LLMs
di: Kim, Junhan, et al.
Pubblicazione: (2026)

Enhancing Instruction Following of LLMs via Activation Steering with Dynamic Rejection
di: Kang, Minjae, et al.
Pubblicazione: (2026)

Incremental Learning of Retrievable Skills For Efficient Continual Task Adaptation
di: Lee, Daehee, et al.
Pubblicazione: (2024)

Merging Triggers, Breaking Backdoors: Defensive Poisoning for Instruction-Tuned Language Models
di: Kim, San, et al.
Pubblicazione: (2026)

Polishing Every Facet of the GEM: Testing Linguistic Competence of LLMs and Humans in Korean
di: Kim, SungHo, et al.
Pubblicazione: (2025)