23171
Admin: @Raminmousa ID: @Machine_learn link: https://t.me/Machine_learn
با عرض سلام اين مقاله دو جايگاه باقي مونده....!
Читать полностью…
🛠️OpenAI just released new guide on how coding agents like GPT-5.1-Codex-Max plug into everyday engineering workflow
📚 Read
@Machine_learn
🔹 Title: TreePO: Bridging the Gap of Policy Optimization and Efficacy and Inference Efficiency with Heuristic Tree-based Modeling
🔹 Publication Date: Published on Aug 24
🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.17445
• PDF: https://arxiv.org/pdf/2508.17445
@Machine_learn
🔹 Title: UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context Learning
🔹 Publication Date: Published on Aug 26
🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.18756
• PDF: https://arxiv.org/pdf/2508.18756
• Github: https://github.com/ZihaoHuang-notabot/Ultra-Sparse-Memory-Network
@Machine_learn
📑 A gentle introduction to pangenomics
📎 Study the paper
@Machine_learn
🔹 Title: UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context Learning
🔹 Publication Date: Published on Aug 26
🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.18756
• PDF: https://arxiv.org/pdf/2508.18756
• Github: https://github.com/ZihaoHuang-notabot/Ultra-Sparse-Memory-Network
@Machine_learn
نفر ۷ از این مقاله پر شد...!
Читать полностью…
State of AI-assisted Software Development
📕 Report
@Machine_learn
The Smol Training Playbook:
The Secrets to Building World-Class LLMs
📚 Read
@Machine_learn
با عرض سلام در حال تنظیم مقاله ای تحت عنوان
Title: MediSeg: Medical Segmentation and classification Recommender system ....
Journal: IEEE Transactions on Medical Imaging
If: 9.8
این کار ۶ ماه طول خواهد کشید و به مسائل بهینه سازی انرژی، جلوگیری از اموزش مجدد شبکه ها، و مسائل تولید کربن در شبکه ها ی عصبی پرداخته خواهد شد.
هزینه مشارکت :
2: 600$
3:500 $
4: 400$
5:300$
6: 200$
7:200$
@Raminmousa
@Machine_learn
@Paper4money
gpt-engineer
gpt-engineer is a project in which you specify what you want in plain English and it iterates to produce a working codebase or scaffolded app. It’s an excellent playground for anyone exploring code-generation agents. And the repo contains clear install/usage instructions and a low-friction dev loop.
Creator: AntonOsika
Stars ⭐️: 55,000
Forked by: 7,300
Github Repo:
https://github.com/AntonOsika/gpt-engineer
@Machine_learn
🔹 Title: OmniHuman-1.5: Instilling an Active Mind in Avatars via Cognitive Simulation
🔹 Publication Date: Published on Aug 26
🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.19209
• PDF: https://arxiv.org/pdf/2508.19209
• Project Page: https://omnihuman-lab.github.io/v1_5/
@Machine_learn
🔹 Title: Understanding Tool-Integrated Reasoning
🔹 Publication Date: Published on Aug 26
🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.19201
• PDF: https://arxiv.org/pdf/2508.19201
• Project Page: https://zhongwenxu.notion.site/Understanding-Tool-Integrated-Reasoning-2551c4e140e3805489fadcc802a1ea83
🔹 Datasets citing this paper:
No datasets found
🔹 Spaces citing this paper:
No spaces found
==================================
@Machine_learn
The LLM Evaluation Framework
🖥 Github: http://github.com/confident-ai/deepeval
📕 Colab: https://colab.research.google.com/drive/1PPxYEBa6eu__LquGoFFJZkhYgWVYE6kh?usp=sharing
🔗 Project: https://deepeval.com
@Machine_learn
🔹 Title: MMTok: Multimodal Coverage Maximization for Efficient Inference of VLMs
🔹 Publication Date: Published on Aug 25
🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.18264
• PDF: https://arxiv.org/pdf/2508.18264
• Project Page: https://project.ironieser.cc/mmtok
@Machine_learn
با عرض سلام برای مقاله زیر نیاز به نفرات ۲ و ۳ داریم.
KG-Psy: A Knowledge-Graph and GPT-5 Based Framework for Personalized Clinical Decision Support in Bipolar Disorder and Borderline Personality Disorder
Abstract: Accurate diagnosis and personalized treatment planning for complex psychiatric disorders such as Bipolar Disorder (BD) and Borderline Personality Disorder (BPD) remain major challenges due to overlapping symptoms, fluctuating mood patterns, and heterogeneous clinical presentations. To address these challenges, we introduce KG-Psy, a hybrid neuro-symbolic framework that combines a domain-specific psychiatric Knowledge Graph (KG) with the advanced reasoning capabilities of GPT-5.
KG-Psy constructs multi-layer psychiatric knowledge graphs encoding symptom trajectories, neural correlates, pharmacological mechanisms, therapeutic guidelines, comorbidities, and behavioral patterns extracted from large-scale clinical literature. GPT-5 is employed to extract clinical entities, infer latent symptom-neural relationships, assess diagnostic likelihoods, and generate patient-specific treatment recommendations. The integration of structured KG reasoning with LLM-based inference allows KG-Psy to produce interpretable, evidence-supported, and clinically actionable outputs.
We evaluated KG-Psy on 310 de-identified psychiatric case reports and 12 expert-validated benchmark scenarios. The framework achieved 91.5% F1-score in distinguishing BD from BPD and an average pathway confidence of 86.9%, indicating robust multi-step inference. In personalized treatment recommendation tasks, KG-Psy achieved 88.7% accuracy, outperforming LLM-only and KG-only baselines by 23% and 31%, respectively.
....
Keywords: Bipolar Disorder, Borderline Personality Disorder, Knowledge Graph, GPT-5, Personalized Treatment
2 :20 milion
3 :15 milion
@Raminmousa
@Machine_learn
@paper4money
🔹 Title: FastMesh:Efficient Artistic Mesh Generation via Component Decoupling
🔹 Publication Date: Published on Aug 26
🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.19188
• PDF: https://arxiv.org/pdf/2508.19188
• Project Page: https://jhkim0759.github.io/projects/FastMesh/
@Machine_learn
Stochastic and deterministic sampling methods in diffusion models produce noticeably different trajectories, but ultimately both reach the same goal.
Diffusion Explorer allows you to visually compare different sampling methods and training objectives of diffusion models by creating visualizations like the one in the 2 videos.
Additionally, you can, for example, train a model on your own dataset and observe how it gradually converges to a sample from the correct distribution.
Check out this GitHub repository:
https://github.com/helblazer811/Diffusion-Explorer
@Machine_learn
Machine Learning Systems
Principles and Practices of Engineering Artificially Intelligent Systems
📚 Read
@Machine_learn
📹 AI in Bioinformatics Overcoming Pitfalls in Statistical, ML and Generative AI Approaches
🎞 Watch
@Machine_learn
🔹 Title: OmniHuman-1.5: Instilling an Active Mind in Avatars via Cognitive Simulation
🔹 Publication Date: Published on Aug 26
🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.19209
• PDF: https://arxiv.org/pdf/2508.19209
• Project Page: https://omnihuman-lab.github.io/v1_5/
@Machine_learn
با عرض سلام در حال تنظیم مقاله ای تحت عنوان
Title: MediSeg: Medical Segmentation and classification Recommender system ....
Journal: IEEE Transactions on Medical Imaging
If: 9.8
این کار ۶ ماه طول خواهد کشید و به مسائل بهینه سازی انرژی، جلوگیری از اموزش مجدد شبکه ها، و مسائل تولید کربن در شبکه ها ی عصبی پرداخته خواهد شد.
هزینه مشارکت :
2: 600$
3:500 $
4: 400$
5:300$
6: 200$
7:200$
@Raminmousa
@Machine_learn
@Paper4money
با عرض سلام دوستاني كه مايل به اين پروژه هستن مي تونن بهمون ملحق بشن
@Raminmousa
5-phase path every ML systems engineer follows but almost no one talks about.
📚 Read
@Machine_learn
A combined UNet++ and LSTM approach for breast ultrasound image segmentation
Author: @Raminmousa
Doi:https://kwnsfk27.r.eu-west-1.awstrack.me/L0/https:%2F%2Fdoi.org%2F10.1016%2Fj.fraope.2025.100385/1/0102019a29ac6163-f3d94222-1d67-4a4b-8bb0-22ed875a5711-000000/VLVCKG-PEYhELOu00YPCeqOuaaA=449
Link:https://www.sciencedirect.com/science/article/pii/S2773186325001732?ref=pdf_download&fr=RR-8&rr=9958aaca19dd11fc
@Machine_learn
دوستانی که نیاز به این مقاله دارن جایگاه ۲، ۳، ۴ خالی می باشد...!
@Raminmousa
با عرض سلام مي خواهيم مقاله اي زیر را ادامه بدیم.
Recurrent Neural Network Basic defiences
نيازمند ٤ نفر هستيم كه بتونن در نگارش و كارها و هزينه كار كمكمون كنند. هزينه نفرات براي اين كار كه ١٨ بنچ مارك باید اجرا بشه. از قرار زير:
1: 700$(❌)
2: 500$✅
3: 400$✅
4: 350$ ✅
دوستاني كه مايل هستن مي تونن به ايدي بنده پيام بدن.
همچنين كدهاي كار، توضيحات و ... در اختيار دوستان قرار ميگيره.
@Raminmousa
@Machine_learn
@Paper4money
🔹 Title: Hermes 4 Technical Report
🔹 Publication Date: Published on Aug 25
🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.18255
• PDF: https://arxiv.org/pdf/2508.18255
• Project Page: https://hermes4.nousresearch.com/
🔹 Datasets citing this paper:
No datasets found
🔹 Spaces citing this paper:
No spaces found
==================================
@Machine_learn
InstructFLIP: Exploring Unified Vision-Language Model for Face Anti-spoofing
🖥 Github: https://github.com/kunkunlin1221/InstructFLIP
📕 Paper: https://arxiv.org/pdf/2507.12060v1.pdf
🔗 Dataset: https://paperswithcode.com/dataset/replay-attack
@Machine_learn
🔹 Title: Limitations of Normalization in Attention Mechanism
🔹 Publication Date: Published on Aug 25
🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.17821
• PDF: https://arxiv.org/pdf/2508.17821
@Machine_learn