Imagen de Portada

Revisiting the Data Sampling in Multimodal Post-training from a Difficulty-Distinguish View

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Qi, Jianyu, Zou, Ding, Yan, Wenrui, Ma, Rui, Li, Jiaxu, Zheng, Zhijie, Yang, Zhiguo, Zhao, Rongchang
Formato:	Preprint
Publicado:	2025
Materias:	Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Reasoning with OmniThought: A Large CoT Dataset with Verbosity and Cognitive Difficulty Annotations
por: Cai, Wenrui, et al.
Publicado: (2025)

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

Learn or Recall? Revisiting Incremental Learning with Pre-trained Language Models
por: Zheng, Junhao, et al.
Publicado: (2023)

AIR: Post-training Data Selection for Reasoning via Attention Head Influence
por: Liu, Jinrui, et al.
Publicado: (2025)

On Predicting the Post-training Potential of Pre-trained LLMs
por: Li, Xiaoyuan, et al.
Publicado: (2026)

Asymmetric Conflict and Synergy in Post-training for LLM-based Multilingual Machine Translation
por: Zheng, Tong, et al.
Publicado: (2025)

Revisiting Generalization Across Difficulty Levels: It's Not So Easy
por: Kordi, Yeganeh, et al.
Publicado: (2025)

Unsupervised Cross-Lingual Part-of-Speech Tagging with Monolingual Corpora Only
por: Zheng, Jianyu
Publicado: (2026)

Effective vocabulary expanding of multilingual language models for extremely low-resource languages
por: Zheng, Jianyu
Publicado: (2026)

On the Impact of Calibration Data in Post-training Quantization and Pruning
por: Williams, Miles, et al.
Publicado: (2023)

Are Transformers in Pre-trained LM A Good ASR Encoder? An Empirical Study
por: An, Keyu, et al.
Publicado: (2024)

Unveiling the Impact of Multimodal Features on Chinese Spelling Correction: From Analysis to Design
por: Zhang, Xiaowu, et al.
Publicado: (2025)

Characterizing Memorization in Diffusion Language Models: Generalized Extraction and Sampling Effects
por: Luo, Xiaoyu, et al.
Publicado: (2026)

Pre-trained Language Models Return Distinguishable Probability Distributions to Unfaithfully Hallucinated Texts
por: Cha, Taehun, et al.
Publicado: (2024)

GM-PRM: A Generative Multimodal Process Reward Model for Multimodal Mathematical Reasoning
por: Zhang, Jianghangfan, et al.
Publicado: (2025)

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

SampleMix: A Sample-wise Pre-training Data Mixing Strategey by Coordinating Data Quality and Diversity
por: Xi, Xiangyu, et al.
Publicado: (2025)

Breaking the Pre-Sampling Barrier: Activation-Informed Difficulty-Aware Self-Consistency
por: Yoon, Taewoong, et al.
Publicado: (2026)

Hierarchical Multimodal Pre-training for Visually Rich Webpage Understanding
por: Xu, Hongshen, et al.
Publicado: (2024)

Got Compute, but No Data: Lessons From Post-training a Finnish LLM
por: Zosa, Elaine, et al.
Publicado: (2025)

Enhancing EEG-to-Text Decoding through Transferable Representations from Pre-trained Contrastive EEG-Text Masked Autoencoder
por: Wang, Jiaqi, et al.
Publicado: (2024)

On Data Synthesis and Post-training for Visual Abstract Reasoning
por: Zhu, Ke, et al.
Publicado: (2025)

DeepDistill: Enhancing LLM Reasoning Capabilities via Large-Scale Difficulty-Graded Data Training
por: Tian, Xiaoyu, et al.
Publicado: (2025)

Improving Similar Case Retrieval Ranking Performance By Revisiting RankSVM
por: Liu, Yuqi, et al.
Publicado: (2025)

Revisiting Classification Taxonomy for Grammatical Errors
por: Zou, Deqing, et al.
Publicado: (2025)

ARES: Multimodal Adaptive Reasoning via Difficulty-Aware Token-Level Entropy Shaping
por: Chen, Shuang, et al.
Publicado: (2025)

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

Pointer-CAD: Unifying B-Rep and Command Sequences via Pointer-based Edges & Faces Selection
por: Qi, Dacheng, et al.
Publicado: (2026)

Rethinking Expert Trajectory Utilization in LLM Post-training for Mathematical Reasoning
por: Ding, Bowen, et al.
Publicado: (2025)

Difficulty-Based Preference Data Selection by DPO Implicit Reward Gap
por: Qi, Xuan, et al.
Publicado: (2025)

RoboUniView: Visual-Language Model with Unified View Representation for Robotic Manipulation
por: Liu, Fanfan, et al.
Publicado: (2024)

Predictive Data Selection: The Data That Predicts Is the Data That Teaches
por: Shum, Kashun, et al.
Publicado: (2025)

ATLAS: A High-Difficulty, Multidisciplinary Benchmark for Frontier Scientific Reasoning
por: Liu, Hongwei, et al.
Publicado: (2025)

Caption First, VQA Second: Knowledge Density, Not Task Format, Drives Multimodal Scaling
por: Zou, Hongjian, et al.
Publicado: (2026)

TimeBill: Time-Budgeted Inference for Large Language Models
por: Fan, Qi, et al.
Publicado: (2025)

Anchoring Bias in Large Language Models: An Experimental Study
por: Lou, Jiaxu, et al.
Publicado: (2024)

Reasoning Core: A Scalable Procedural Data Generation Suite for Symbolic Pre-training and Post-Training
por: Lacombe, Valentin, et al.
Publicado: (2026)

Text Difficulty Study: Do machines behave the same as humans regarding text difficulty?
por: Chen, Bowen, et al.
Publicado: (2022)

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

DataProphet: Demystifying Supervision Data Generalization in Multimodal LLMs
por: Qi, Xuan, et al.
Publicado: (2026)