ML Research Hub
32.8K subscribers
4.34K photos
264 videos
23 files
4.68K links
Advancing research in Machine Learning – practical insights, tools, and techniques for researchers.

Admin: @HusseinSheikho || @Hussein_Sheikho
Download Telegram
πŸ”Ή Title: From Charts to Code: A Hierarchical Benchmark for Multimodal Models

πŸ”Ή Publication Date: Published on Oct 20

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.17932
β€’ PDF: https://arxiv.org/pdf/2510.17932
β€’ Project Page: https://csu-jpg.github.io/Chart2Code.github.io/
β€’ Github: https://github.com/CSU-JPG/Chart2Code

πŸ”Ή Datasets citing this paper:
β€’ https://huggingface.co/datasets/CSU-JPG/Chart2Code

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: Attention Sinks in Diffusion Language Models

πŸ”Ή Publication Date: Published on Oct 17

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.15731
β€’ PDF: https://arxiv.org/pdf/2510.15731

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: Directional Reasoning Injection for Fine-Tuning MLLMs

πŸ”Ή Publication Date: Published on Oct 16

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.15050
β€’ PDF: https://arxiv.org/pdf/2510.15050
β€’ Github: https://github.com/WikiChao/DRIFT

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: See the Text: From Tokenization to Visual Reading

πŸ”Ή Publication Date: Published on Oct 21

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.18840
β€’ PDF: https://arxiv.org/pdf/2510.18840

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
❀1
πŸ”Ή Title: DeLeaker: Dynamic Inference-Time Reweighting For Semantic Leakage Mitigation in Text-to-Image Models

πŸ”Ή Publication Date: Published on Oct 16

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.15015
β€’ PDF: https://arxiv.org/pdf/2510.15015

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: Language Models are Injective and Hence Invertible

πŸ”Ή Publication Date: Published on Oct 17

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.15511
β€’ PDF: https://arxiv.org/pdf/2510.15511

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
❀1
πŸ”Ή Title: Decomposed Attention Fusion in MLLMs for Training-Free Video Reasoning Segmentation

πŸ”Ή Publication Date: Published on Oct 22

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.19592
β€’ PDF: https://arxiv.org/pdf/2510.19592
β€’ Project Page: https://www.jshyun.me/projects/decaf
β€’ Github: https://github.com/HYUNJS/DecAF

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ€–πŸ§  LangGraph by LangChain-AI: The Framework Powering Stateful, Long-Running AI Agents

πŸ—“οΈ 23 Oct 2025
πŸ“š AI News & Trends

As artificial intelligence continues to reshape industries, one major challenge remains building reliable, stateful and long-running AI agents that can handle complex workflows over time. Traditional AI frameworks often focus on short interactions, lacking the infrastructure to manage persistent states, durable memory or human feedback loops. That’s where LangGraph from LangChain-AI steps in. Trusted by ...

#LangGraph #LangChainAI #AIAgents #StatefulAI #LongRunningAgents #ArtificialIntelligence
πŸ€–πŸ§  Master Machine Learning: Explore the Ultimate β€œMachine-Learning-Tutorials” Repository

πŸ—“οΈ 23 Oct 2025
πŸ“š AI News & Trends

In today’s data-driven world, Machine Learning (ML) has become the cornerstone of modern technology from intelligent chatbots to predictive analytics and recommendation systems. However, mastering ML isn’t just about coding, it requires a structured understanding of algorithms, statistics, optimization techniques and real-world problem-solving. That’s where Ujjwal Karn’s Machine-Learning-Tutorials GitHub repository stands out. This open-source, topic-wise ...

#MachineLearning #MLTutorials #ArtificialIntelligence #DataScience #OpenSource #AIEducation
πŸ”Ή Title: What Questions Should Robots Be Able to Answer? A Dataset of User Questions for Explainable Robotics

πŸ”Ή Publication Date: Published on Oct 18

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.16435
β€’ PDF: https://arxiv.org/pdf/2510.16435

πŸ”Ή Datasets citing this paper:
β€’ https://huggingface.co/datasets/lwachowiak/xai-questions-dataset

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: Steering Autoregressive Music Generation with Recursive Feature Machines

πŸ”Ή Publication Date: Published on Oct 21

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.19127
β€’ PDF: https://arxiv.org/pdf/2510.19127
β€’ Github: https://github.com/astradzhao/music-rfm

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: SAVANT: Semantic Analysis with Vision-Augmented Anomaly deTection

πŸ”Ή Publication Date: Published on Oct 20

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.18034
β€’ PDF: https://arxiv.org/pdf/2510.18034

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: Accelerating Vision Transformers with Adaptive Patch Sizes

πŸ”Ή Publication Date: Published on Oct 20

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.18091
β€’ PDF: https://arxiv.org/pdf/2510.18091
β€’ Github: https://github.com/rccchoudhury/apt

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: Text or Pixels? It Takes Half: On the Token Efficiency of Visual Text Inputs in Multimodal LLMs

πŸ”Ή Publication Date: Published on Oct 21

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.18279
β€’ PDF: https://arxiv.org/pdf/2510.18279
β€’ Github: https://github.com/yanhong-lbh/text_or_pixels

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: Every Question Has Its Own Value: Reinforcement Learning with Explicit Human Values

πŸ”Ή Publication Date: Published on Oct 23

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.20187
β€’ PDF: https://arxiv.org/pdf/2510.20187

πŸ”Ή Datasets citing this paper:
β€’ https://huggingface.co/datasets/sarosavo/RLEV

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: HoloCine: Holistic Generation of Cinematic Multi-Shot Long Video Narratives

πŸ”Ή Publication Date: Published on Oct 23

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.20822
β€’ PDF: https://arxiv.org/pdf/2510.20822
β€’ Project Page: https://holo-cine.github.io

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: LayerComposer: Interactive Personalized T2I via Spatially-Aware Layered Canvas

πŸ”Ή Publication Date: Published on Oct 23

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.20820
β€’ PDF: https://arxiv.org/pdf/2510.20820
β€’ Project Page: https://snap-research.github.io/layercomposer/

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: AlphaFlow: Understanding and Improving MeanFlow Models

πŸ”Ή Publication Date: Published on Oct 23

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.20771
β€’ PDF: https://arxiv.org/pdf/2510.20771
β€’ Github: https://github.com/snap-research/alphaflow

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: Open-o3 Video: Grounded Video Reasoning with Explicit Spatio-Temporal Evidence

πŸ”Ή Publication Date: Published on Oct 23

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.20579
β€’ PDF: https://arxiv.org/pdf/2510.20579
β€’ Project Page: https://marinero4972.github.io/projects/Open-o3-Video/

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: Loopholing Discrete Diffusion: Deterministic Bypass of the Sampling Wall

πŸ”Ή Publication Date: Published on Oct 22

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.19304
β€’ PDF: https://arxiv.org/pdf/2510.19304

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT
πŸ”Ή Title: ImpossibleBench: Measuring LLMs' Propensity of Exploiting Test Cases

πŸ”Ή Publication Date: Published on Oct 23

πŸ”Ή Paper Links:
β€’ arXiv Page: https://arxiv.org/abs/2510.20270
β€’ PDF: https://arxiv.org/pdf/2510.20270

πŸ”Ή Datasets citing this paper:
No datasets found

πŸ”Ή Spaces citing this paper:
No spaces found
==================================

For more data science resources:
βœ“ https://t.me/DataScienceT