ML Research Hub – Telegram
ML Research Hub
32.7K subscribers
4.09K photos
237 videos
23 files
4.41K links
Advancing research in Machine Learning – practical insights, tools, and techniques for researchers.

Admin: @HusseinSheikho || @Hussein_Sheikho
Download Telegram
🔹 Title: MultiEdit: Advancing Instruction-based Image Editing on Diverse and Challenging Tasks

🔹 Publication Date: Published on Sep 18

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.14638
• PDF: https://arxiv.org/pdf/2509.14638
• Project Page: https://huggingface.co/datasets/inclusionAI/MultiEdit

🔹 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: RynnVLA-001: Using Human Demonstrations to Improve Robot Manipulation

🔹 Publication Date: Published on Sep 18

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.15212
• PDF: https://arxiv.org/pdf/2509.15212
• Github: https://github.com/alibaba-damo-academy/RynnVLA-001

🔹 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: WorldForge: Unlocking Emergent 3D/4D Generation in Video Diffusion Model via Training-Free Guidance

🔹 Publication Date: Published on Sep 18

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.15130
• PDF: https://arxiv.org/pdf/2509.15130
• Project Page: https://worldforge-agi.github.io/

🔹 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: AToken: A Unified Tokenizer for Vision

🔹 Publication Date: Published on Sep 17

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.14476
• PDF: https://arxiv.org/pdf/2509.14476

🔹 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: Understand Before You Generate: Self-Guided Training for Autoregressive Image Generation

🔹 Publication Date: Published on Sep 18

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.15185
• PDF: https://arxiv.org/pdf/2509.15185

🔹 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: Unleashing the Potential of Multimodal LLMs for Zero-Shot Spatio-Temporal Video Grounding

🔹 Publication Date: Published on Sep 18

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.15178
• PDF: https://arxiv.org/pdf/2509.15178
• Github: https://github.com/zaiquanyang/LLaVA_Next_STVG

🔹 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: FSG-Net: Frequency-Spatial Synergistic Gated Network for High-Resolution Remote Sensing Change Detection

🔹 Publication Date: Published on Sep 8

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.06482
• PDF: https://arxiv.org/pdf/2509.06482
• Github: https://github.com/zxXie-Air/FSG-Net

🔹 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: Reasoning over Boundaries: Enhancing Specification Alignment via Test-time Delibration

🔹 Publication Date: Published on Sep 18

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.14760
• PDF: https://arxiv.org/pdf/2509.14760

🔹 Datasets citing this paper:
https://huggingface.co/datasets/zzzhr97/SpecBench

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

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: EchoVLM: Dynamic Mixture-of-Experts Vision-Language Model for Universal Ultrasound Intelligence

🔹 Publication Date: Published on Sep 18

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.14977
• PDF: https://arxiv.org/pdf/2509.14977
• Github: https://github.com/Asunatan/EchoVLM

🔹 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
🥳Working on web scraping, data automation, or machine learning projects?
RapidProxy provides developers with a reliable proxy network built for scale.
🌐 https://www.rapidproxy.io/?ref=tst

Why Developers Choose RapidProxy

🔡 Traffic Never Expires
No monthly lock-in. Pay once, use when you need—perfect for long-term projects.

🔡 Free Trial for Testing
Evaluate performance before committing. Ideal for experimenting with your Python workflows.

🔡 Large Residential IP Pool
90M+ clean residential IPs across 200+ countries and cities. Great for bypassing restrictions, collecting data, or managing accounts.

🔡 HTTP(S) /SOCKS5 support
HTTP(S) & SOCKS5 compatibility ensures smooth integration with tools like requests, Scrapy, or Selenium.

💵Pricing

Residential Proxy: from $0.65/GB, traffic never expires.
Static (ISP) Residential Proxy: $5/IP, unlimited traffic for 30 days (easy renewal).

Get started in minutes:
1️⃣ Sign up at RapidProxy.io
2️⃣ Claim your free trial traffic
3️⃣ Scale your scraping, automation, or research projects with confidence

📞Contact Us

✈️Telegram: @rapidproxyio
📱WhatsApp: Rapidproxy.io
💌Email: support@rapidproxy.io / helena@rapidproxy.io

⚡️Performance You Can Rely On

☑️99.9% uptime guarantee
☑️Ultra-low 0.38s average response time
☑️Unlimited concurrent sessions for high-volume tasks

RapidProxy — Fast · Reliable · Developer-Friendly
Your trusted partner for Python scraping and global online operations.
Please open Telegram to view this post
VIEW IN TELEGRAM
3
🔹 Title: Mind the Gap: A Closer Look at Tokenization for Multiple-Choice Question Answering with LLMs

🔹 Publication Date: Published on Sep 18

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.15020
• PDF: https://arxiv.org/pdf/2509.15020

🔹 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: RecoWorld: Building Simulated Environments for Agentic Recommender Systems

🔹 Publication Date: Published on Sep 12

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.10397
• PDF: https://arxiv.org/pdf/2509.10397

🔹 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: Agentic Software Engineering: Foundational Pillars and a Research Roadmap

🔹 Publication Date: Published on Sep 7

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.06216
• PDF: https://arxiv.org/pdf/2509.06216

🔹 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: Developer-LLM Conversations: An Empirical Study of Interactions and Generated Code Quality

🔹 Publication Date: Published on Sep 12

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.10402
• PDF: https://arxiv.org/pdf/2509.10402

🔹 Datasets citing this paper:
https://huggingface.co/datasets/Suzhen/CodeChat

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

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: Apertus: Democratizing Open and Compliant LLMs for Global Language Environments

🔹 Publication Date: Published on Sep 17

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.14233
• PDF: https://arxiv.org/pdf/2509.14233
• Project Page: https://www.swiss-ai.org/apertus
• Github: https://github.com/swiss-ai/apertus-tech-report

🔹 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: Can Multimodal LLMs See Materials Clearly? A Multimodal Benchmark on Materials Characterization

🔹 Publication Date: Published on Sep 11

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.09307
• PDF: https://arxiv.org/pdf/2509.09307

🔹 Datasets citing this paper:
https://huggingface.co/datasets/FreedomIntelligence/MatCha

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

For more data science resources:
https://news.1rj.ru/str/DataScienceT
🔹 Title: EdiVal-Agent: An Object-Centric Framework for Automated, Scalable, Fine-Grained Evaluation of Multi-Turn Editing

🔹 Publication Date: Published on Sep 16

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2509.13399
• PDF: https://arxiv.org/pdf/2509.13399
• Github: https://tianyucodings.github.io/EdiVAL-page/

🔹 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
🌟 Swiftproxy – Fast, Secure, and Reliable Proxies for Every Project

Looking for proxies that are stable, high-speed, and ready for any online task? Swiftproxy is your go-to solution! From web scraping and data analytics to automation and competitive research, our proxies help you work smarter, faster, and without limits.
🌐 https://www.swiftproxy.net/?ref=python53

Why Swiftproxy Stands Out:

🔄 Rotating & Sticky IPs

Switch IPs seamlessly or maintain the same one for longer sessions. Avoid blocks, CAPTCHAs, and interruptions effortlessly.

🌍 Global Coverage & Advanced Targeting
Access over 90 million IPs across 220+ countries. Pinpoint IPs by country, city, or ASN for precise data collection.

⚡️ Fast, Secure, & Reliable

Enjoy high-speed servers with 99.8% uptime and advanced encryption, ensuring seamless and secure data collection.

💻 Flexible & Compatible
Supports HTTP(S) and SOCKS5 protocols and integrates easily with over 650 tools and applications.

Flexible Data Usage
Residential proxies: Non-expiring bandwidth starting from just $0.7/GB — use it whenever you need.
Static residential proxies: Valid for 30 days, priced from $6/IP, giving you maximum control.

🔥Kickstart your projects with Swiftproxy’s reliable residential proxies and make your workflow smoother than ever!
👉 https://www.swiftproxy.net/?ref=python53

🎯 Get Started Quickly:
1. Register a new account on Swiftproxy
2. Contact support to claim 500MB free residential proxy traffic
🎊Use exclusive code SWIFT90 for 10% off your first purchase!

💫Join Swiftproxy Now: https://news.1rj.ru/str/swiftproxynetofficial
5