ML Research Hub – Telegram
ML Research Hub
32.8K subscribers
4.41K photos
272 videos
23 files
4.77K links
Advancing research in Machine Learning – practical insights, tools, and techniques for researchers.

Admin: @HusseinSheikho || @Hussein_Sheikho
Download Telegram
🔹 Title: SQUARE: Semantic Query-Augmented Fusion and Efficient Batch Reranking for Training-free Zero-Shot Composed Image Retrieval

🔹 Publication Date: Published on Sep 30

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.26330
• PDF: https://arxiv.org/pdf/2509.26330

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: RLAD: Training LLMs to Discover Abstractions for Solving Reasoning Problems

🔹 Publication Date: Published on Oct 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.02263
• PDF: https://arxiv.org/pdf/2510.02263

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Generalized Parallel Scaling with Interdependent Generations

🔹 Publication Date: Published on Oct 1

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.01143
• PDF: https://arxiv.org/pdf/2510.01143

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔥1
🔹 Title: The Rogue Scalpel: Activation Steering Compromises LLM Safety

🔹 Publication Date: Published on Sep 26

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.22067
• PDF: https://arxiv.org/pdf/2509.22067

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: IoT-MCP: Bridging LLMs and IoT Systems Through Model Context Protocol

🔹 Publication Date: Published on Sep 25

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.01260
• PDF: https://arxiv.org/pdf/2510.01260

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: ModernVBERT: Towards Smaller Visual Document Retrievers

🔹 Publication Date: Published on Oct 1

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.01149
• PDF: https://arxiv.org/pdf/2510.01149
• Project Page: https://huggingface.co/ModernVBERT

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Optimizing What Matters: AUC-Driven Learning for Robust Neural Retrieval

🔹 Publication Date: Published on Sep 30

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.00137
• PDF: https://arxiv.org/pdf/2510.00137

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: One-Token Rollout: Guiding Supervised Fine-Tuning of LLMs with Policy Gradient

🔹 Publication Date: Published on Sep 30

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.26313
• PDF: https://arxiv.org/pdf/2509.26313

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Drawing Conclusions from Draws: Rethinking Preference Semantics in Arena-Style LLM Evaluation

🔹 Publication Date: Published on Oct 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.02306
• PDF: https://arxiv.org/pdf/2510.02306
• Github: https://github.com/daemon/lmarena-draws

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Think Right: Learning to Mitigate Under-Over Thinking via Adaptive, Attentive Compression

🔹 Publication Date: Published on Oct 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.01581
• PDF: https://arxiv.org/pdf/2510.01581
• Github: https://github.com/joykirat18/TRAAC

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Controlled Generation for Private Synthetic Text

🔹 Publication Date: Published on Sep 30

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.25729
• PDF: https://arxiv.org/pdf/2509.25729

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
1
🔹 Title: Sparse Query Attention (SQA): A Computationally Efficient Attention Mechanism with Query Heads Reduction

🔹 Publication Date: Published on Oct 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.01817
• PDF: https://arxiv.org/pdf/2510.01817
• Project Page: https://rxai.dev
• Github: https://github.com/RxAI-dev/rxnn-attention

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Aristotle: IMO-level Automated Theorem Proving

🔹 Publication Date: Published on Oct 1

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.01346
• PDF: https://arxiv.org/pdf/2510.01346

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Fine-Grained Detection of Context-Grounded Hallucinations Using LLMs

🔹 Publication Date: Published on Sep 26

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.22582
• PDF: https://arxiv.org/pdf/2509.22582

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Rethinking Thinking Tokens: LLMs as Improvement Operators

🔹 Publication Date: Published on Oct 1

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.01123
• PDF: https://arxiv.org/pdf/2510.01123

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
Forwarded from Machine Learning
📌 MobileNetV2 Paper Walkthrough: The Smarter Tiny Giant

🗂 Category: DEEP LEARNING

🕒 Date: 2025-10-03 | ⏱️ Read time: 28 min read

Understanding and implementing MobileNetV2 with PyTorch  — the next generation of MobileNetV1
1
🔹 Title: Improving GUI Grounding with Explicit Position-to-Coordinate Mapping

🔹 Publication Date: Published on Oct 3

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.03230
• PDF: https://arxiv.org/pdf/2510.03230

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Self-Improvement in Multimodal Large Language Models: A Survey

🔹 Publication Date: Published on Oct 3

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.02665
• PDF: https://arxiv.org/pdf/2510.02665

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: How Confident are Video Models? Empowering Video Models to Express their Uncertainty

🔹 Publication Date: Published on Oct 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.02571
• PDF: https://arxiv.org/pdf/2510.02571
• Project Page: https://s-qubed.github.io/
• Github: https://github.com/irom-princeton/s-qubed

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Compose Your Policies! Improving Diffusion-based or Flow-based Robot Policies via Test-time Distribution-level Composition

🔹 Publication Date: Published on Oct 1

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.01068
• PDF: https://arxiv.org/pdf/2510.01068
• Project Page: https://sagecao1125.github.io/GPC-Site/
• Github: https://sagecao1125.github.io/GPC-Site/

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT
1
🔹 Title: REPAIR: Robust Editing via Progressive Adaptive Intervention and Reintegration

🔹 Publication Date: Published on Oct 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.01879
• PDF: https://arxiv.org/pdf/2510.01879

🔹 Datasets citing this paper:
No datasets found

🔹 Spaces citing this paper:
No spaces found
==================================

For more data science resources:
https://news.1rj.ru/str/DataScienceT