文件夹: Papers/LLM--and--VLM/Pretraining--and--Architecture

2026年5月

From Pixels to Words -- Towards Native One-Vision Models at Scale

2026年5月

Lance: Unified Multimodal Modeling by Multi-Task Synergy

2026年5月

LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning

2026年5月

LLaVA-OneVision-2: Towards Next-Generation Perceptual Intelligence

2026年5月

LocateAnything: Fast and High-Quality Vision-Language Grounding with Parallel Box Decoding

2026年5月

See What I Mean: Aligning Vision and Language Representations for Video Fine-grained Object Understanding

2026年5月

Toward Native Multimodal Modeling: A Roadmap

2026年5月

Understanding Self-Supervised Learning via Latent Distribution Matching

2026年4月

Adam's Law: Textual Frequency Law on Large Language Models

2026年4月

MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Modal Interaction

2026年3月

DataFlex: A Unified Framework for Data-Centric Dynamic Training of Large Language Models

2026年3月

InternVL-U: Democratizing Unified Multimodal Models for Understanding, Reasoning, Generation and Editing

2026年3月

Penguin-VL: Exploring the Efficiency Limits of VLM with LLM-based Vision Encoders

2026年2月

Semantic Tube Prediction: Beating LLM Data Efficiency with JEPA

2026年1月

Soft Clustering Anchors for Self-Supervised Speech Representation Learning in Joint Embedding Prediction Architectures

2025年12月

JEPA as a Neural Tokenizer: Learning Robust Speech Representations with Density Adaptive Attention

2025年12月

VL-JEPA: Joint Embedding Predictive Architecture for Vision-language

2025年11月

Qwen3-VL Technical Report

2025年9月

LLM-JEPA: Large Language Models Meet Joint Embedding Predictive Architectures

2025年9月

Qwen3-Omni Technical Report

2025年8月

GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models

2025年8月

InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency

2025年7月

Paper Notes

探索

文件夹: Papers/LLM--and--VLM/Pretraining--and--Architecture

From Pixels to Words -- Towards Native One-Vision Models at Scale

Lance: Unified Multimodal Modeling by Multi-Task Synergy

LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning

LLaVA-OneVision-2: Towards Next-Generation Perceptual Intelligence

LocateAnything: Fast and High-Quality Vision-Language Grounding with Parallel Box Decoding

See What I Mean: Aligning Vision and Language Representations for Video Fine-grained Object Understanding

Toward Native Multimodal Modeling: A Roadmap

Understanding Self-Supervised Learning via Latent Distribution Matching

Adam's Law: Textual Frequency Law on Large Language Models

MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Modal Interaction

DataFlex: A Unified Framework for Data-Centric Dynamic Training of Large Language Models

InternVL-U: Democratizing Unified Multimodal Models for Understanding, Reasoning, Generation and Editing

Penguin-VL: Exploring the Efficiency Limits of VLM with LLM-based Vision Encoders

Semantic Tube Prediction: Beating LLM Data Efficiency with JEPA

Soft Clustering Anchors for Self-Supervised Speech Representation Learning in Joint Embedding Prediction Architectures

JEPA as a Neural Tokenizer: Learning Robust Speech Representations with Density Adaptive Attention

VL-JEPA: Joint Embedding Predictive Architecture for Vision-language

Qwen3-VL Technical Report

LLM-JEPA: Large Language Models Meet Joint Embedding Predictive Architectures

Qwen3-Omni Technical Report

GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models

InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency

Kimi K2: Open Agentic Intelligence