✨LSRIF: Logic-Structured Reinforcement Learning for Instruction Following

📝 Summary:
A logic-structured training framework explicitly models instruction logic through constraint-aware reward mechanisms, improving instruction-following and reasoning capabilities in large language model...

🔹 Publication Date: Published on Jan 10

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2601.06431
• PDF: https://arxiv.org/pdf/2601.06431

==================================

For more data science resources:
✓ https://t.iss.one/DataScienceT

#AI #DataScience #MachineLearning #HuggingFace #Research

✨TAG-MoE: Task-Aware Gating for Unified Generative Mixture-of-Experts

📝 Summary:
A novel framework injects semantic intent into Mixture-of-Experts routing for image generation and editing, resolving task interference through hierarchical task annotation and predictive alignment re...

🔹 Publication Date: Published on Jan 12

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2601.08881
• PDF: https://arxiv.org/pdf/2601.08881
• Project Page: https://yuci-gpt.github.io/TAG-MoE/

==================================

For more data science resources:
✓ https://t.iss.one/DataScienceT

#AI #DataScience #MachineLearning #HuggingFace #Research

✨WildRayZer: Self-supervised Large View Synthesis in Dynamic Environments

📝 Summary:
WildRayZer is a self-supervised framework for novel view synthesis in dynamic environments that uses analysis-by-synthesis to handle moving cameras and objects through motion masking and gradient gati...

🔹 Publication Date: Published on Jan 15

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2601.10716
• PDF: https://arxiv.org/pdf/2601.10716

==================================

For more data science resources:
✓ https://t.iss.one/DataScienceT

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Toward Ultra-Long-Horizon Agentic Science: Cognitive Accumulation for Machine Learning Engineering

📝 Summary:
ML-Master 2.0 enables ultra-long-horizon AI autonomy for machine learning engineering. It uses Hierarchical Cognitive Caching to accumulate knowledge from execution, decoupling short-term actions from long-term strategy, achieving state-of-the-art results.

🔹 Publication Date: Published on Jan 15

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2601.10402
• PDF: https://arxiv.org/pdf/2601.10402
• Project Page: https://sjtu-sai-agents.github.io/ML-Master/
• Github: https://github.com/sjtu-sai-agents/ML-Master

==================================

For more data science resources:
✓ https://t.iss.one/DataScienceT

#AI #MachineLearning #AutonomousAI #AIAgents #CognitiveAI

✨Memory Bank Compression for Continual Adaptation of Large Language Models

📝 Summary:
Memory-augmented continual learning for LLMs faces growing memory bank issues. MBC compresses these banks via codebook optimization and an online resetting mechanism, using Key-Value Low-Rank Adaptation. It reduces bank size to 0.3 percent while maintaining high accuracy.

🔹 Publication Date: Published on Jan 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2601.00756
• PDF: https://arxiv.org/pdf/2601.00756
• Github: https://github.com/Thomkat/MBC

==================================

For more data science resources:
✓ https://t.iss.one/DataScienceT

#LLMs #ContinualLearning #MemoryCompression #MachineLearning #DeepLearning

✨Patient-Similarity Cohort Reasoning in Clinical Text-to-SQL

📝 Summary:
CLINSQL is a new benchmark for evaluating text-to-SQL models on complex clinical tasks, including patient similarity, using real EHR data. Current models achieve moderate execution scores but remain far from clinical reliability for real-world EHR analytics.

🔹 Publication Date: Published on Jan 14

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2601.09876
• PDF: https://arxiv.org/pdf/2601.09876
• Github: https://github.com/Barryshen1/ClinSQL

✨ Datasets citing this paper:
• https://huggingface.co/datasets/yifeis02/ClinSQL

==================================

For more data science resources:
✓ https://t.iss.one/DataScienceT

#AI #DataScience #MachineLearning #HuggingFace #Research

✨V-DPM: 4D Video Reconstruction with Dynamic Point Maps

📝 Summary:
Dynamic Point Maps extended to video input through V-DPM framework achieve state-of-the-art 3D and 4D reconstruction by recovering both dynamic depth and full 3D motion of scene points. AI-generated s...

🔹 Publication Date: Published on Jan 14

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2601.09499
• PDF: https://arxiv.org/pdf/2601.09499
• Project Page: https://www.robots.ox.ac.uk/~vgg/research/vdpm/
• Github: https://github.com/eldar/vdpm

==================================

For more data science resources:
✓ https://t.iss.one/DataScienceT

#AI #DataScience #MachineLearning #HuggingFace #Research

✨PACEvolve: Enabling Long-Horizon Progress-Aware Consistent Evolution

📝 Summary:
PACEvolve framework addresses key failure modes in LLM evolutionary search through hierarchical context management, momentum-based backtracking, and adaptive sampling policies for improved self-improv...

🔹 Publication Date: Published on Jan 15

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2601.10657
• PDF: https://arxiv.org/pdf/2601.10657

==================================

For more data science resources:
✓ https://t.iss.one/DataScienceT

#AI #DataScience #MachineLearning #HuggingFace #Research

✨RigMo: Unifying Rig and Motion Learning for Generative Animation

📝 Summary:
RigMo unifies rig and motion learning directly from raw mesh sequences, encoding deformations into compact latent spaces. This framework generates interpretable, plausible 3D animation, offering superior reconstruction and generalization over baselines.

🔹 Publication Date: Published on Jan 10

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2601.06378
• PDF: https://arxiv.org/pdf/2601.06378
• Project Page: https://rigmo-page.github.io/
• Github: https://rigmo-page.github.io

==================================

For more data science resources:
✓ https://t.iss.one/DataScienceT

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Demystifying the Slash Pattern in Attention: The Role of RoPE

📝 Summary:
Slash-Dominant Heads in LLMs emerge when queries and keys are almost rank-one and Rotary Position Embedding has dominant medium-high frequencies. Theoretical proof shows these conditions, combined with gradient descent, explain their emergence.

🔹 Publication Date: Published on Jan 13

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2601.08297
• PDF: https://arxiv.org/pdf/2601.08297

==================================

For more data science resources:
✓ https://t.iss.one/DataScienceT

#AI #DataScience #MachineLearning #HuggingFace #Research