Imagen de Portada

Breaking the Pre-Sampling Barrier: Activation-Informed Difficulty-Aware Self-Consistency

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Yoon, Taewoong, Jeong, Geunyeong, Park, Geon, Yeom, Sihyeong, Kim, Harksoo
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

STEAM: A Semantic-Level Knowledge Editing Framework for Large Language Models
por: Jeong, Geunyeong, et al.
Publicado: (2025)

Exploring the Impact of Instruction-Tuning on LLM's Susceptibility to Misinformation
por: Han, Kyubeen, et al.
Publicado: (2025)

Exploring the Trade-Offs: Quantization Methods, Task Difficulty, and Model Size in Large Language Models From Edge to Giant
por: Lee, Jemin, et al.
Publicado: (2024)

Can Large Language Models Differentiate Harmful from Argumentative Essays? Steps Toward Ethical Essay Scoring
por: Kim, Hongjin, et al.
Publicado: (2026)

Generation-Based and Emotion-Reflected Memory Update: Creating the KEEM Dataset for Better Long-Term Conversation
por: Kang, Jeonghyun, et al.
Publicado: (2026)

Reliability-Aware Adaptive Self-Consistency for Efficient Sampling in LLM Reasoning
por: Kim, Junseok, et al.
Publicado: (2026)

Breaking Language Barriers: Cross-Lingual Continual Pre-Training at Scale
por: Zheng, Wenzhen, et al.
Publicado: (2024)

DAST: Difficulty-Aware Self-Training on Large Language Models
por: Xue, Boyang, et al.
Publicado: (2025)

A Survey on Inference Engines for Large Language Models: Perspectives on Optimization and Efficiency
por: Park, Sihyeong, et al.
Publicado: (2025)

ToDi: Token-wise Distillation via Fine-Grained Divergence Control
por: Jung, Seongryong, et al.
Publicado: (2025)

Garbage Attention in Large Language Models: BOS Sink Heads and Sink-aware Pruning
por: Sok, Jaewon, et al.
Publicado: (2026)

EpiCaR: Knowing What You Don't Know Matters for Better Reasoning in LLMs
por: Yeom, Jewon, et al.
Publicado: (2026)

Hallucination as Commitment Failure: Larger LLMs Misfire Despite Knowing the Answer
por: Yeom, Jewon, et al.
Publicado: (2026)

Learning When to Sample: Confidence-Aware Self-Consistency for Efficient LLM Chain-of-Thought Reasoning
por: Xiong, Juming, et al.
Publicado: (2026)

EdiText: Controllable Coarse-to-Fine Text Editing with Diffusion Language Models
por: Lee, Che Hyun, et al.
Publicado: (2025)

Reasoning Aware Self-Consistency: Leveraging Reasoning Paths for Efficient LLM Sampling
por: Wan, Guangya, et al.
Publicado: (2024)

Make Every Penny Count: Difficulty-Adaptive Self-Consistency for Cost-Efficient Reasoning
por: Wang, Xinglin, et al.
Publicado: (2024)

Linguistically Informed Graph Model and Semantic Contrastive Learning for Korean Short Text Classification
por: Yoo, JaeGeon, et al.
Publicado: (2026)

Revisiting Self-Play Preference Optimization: On the Role of Prompt Difficulty
por: Xiao, Yao, et al.
Publicado: (2025)

Breaking the Script Barrier in Multilingual Pre-Trained Language Models with Transliteration-Based Post-Training Alignment
por: Xhelili, Orgest, et al.
Publicado: (2024)

Theme-Explanation Structure for Table Summarization using Large Language Models: A Case Study on Korean Tabular Data
por: Kwack, TaeYoon, et al.
Publicado: (2025)

DARO: Difficulty-Aware Reweighting Policy Optimization
por: Zhou, Jingyu, et al.
Publicado: (2025)

Boundless Byte Pair Encoding: Breaking the Pre-tokenization Barrier
por: Schmidt, Craig W., et al.
Publicado: (2025)

Breaking the Language Barrier: Can Direct Inference Outperform Pre-Translation in Multilingual LLM Applications?
por: Intrator, Yotam, et al.
Publicado: (2024)

HeaPA: Difficulty-Aware Heap Sampling and On-Policy Query Augmentation for LLM Reinforcement Learning
por: Wang, Weiqi, et al.
Publicado: (2026)

CODA: Difficulty-Aware Compute Allocation for Adaptive Reasoning
por: Wu, Siye, et al.
Publicado: (2026)

Online Difficulty Filtering for Reasoning Oriented Reinforcement Learning
por: Bae, Sanghwan, et al.
Publicado: (2025)

SDS KoPub VDR: A Benchmark Dataset for Visual Document Retrieval in Korean Public Documents
por: Lee, Jaehoon, et al.
Publicado: (2025)

Typed-RAG: Type-Aware Decomposition of Non-Factoid Questions for Retrieval-Augmented Generation
por: Lee, DongGeon, et al.
Publicado: (2025)

Guard Vector: Beyond English LLM Guardrails with Task-Vector Composition and Streaming-Aware Prefix SFT
por: Lee, Wonhyuk, et al.
Publicado: (2025)

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

Enhancing Time Awareness in Generative Recommendation
por: Lee, Sunkyung, et al.
Publicado: (2025)

PACE: Prefix-Protected and Difficulty-Aware Compression for Efficient Reasoning
por: Feng, Ruixiang, et al.
Publicado: (2026)

Where Reasoning Breaks: Logic-Aware Path Selection by Controlling Logical Connectives in LLMs Reasoning Chains
por: Park, Seunghyun, et al.
Publicado: (2026)

Latent Self-Consistency for Reliable Majority-Set Selection in Short- and Long-Answer Reasoning
por: Oh, Jungsuk, et al.
Publicado: (2025)

Are Any-to-Any Models More Consistent Across Modality Transfers Than Specialists?
por: Chung, Jiwan, et al.
Publicado: (2025)

From Dense to Dynamic: Token-Difficulty Driven MoEfication of Pre-Trained LLMs
por: Nishu, Kumari, et al.
Publicado: (2025)

Mirror Speculative Decoding: Breaking the Serial Barrier in LLM Inference
por: Bhendawade, Nikhil, et al.
Publicado: (2025)

Everyday Physics in Korean Contexts: A Culturally Grounded Physical Reasoning Benchmark
por: Jeong, Jihae, et al.
Publicado: (2025)

BottleHumor: Self-Informed Humor Explanation using the Information Bottleneck Principle
por: Hwang, EunJeong, et al.
Publicado: (2025)