✨OneVision-Encoder: Codec-Aligned Sparsity as a Foundational Principle for Multimodal Intelligence

📝 Summary:
OneVision-Encoder improves visual understanding by aligning architectures with video compression principles. It uses codec-aligned sparsity to focus on high-entropy regions, significantly boosting efficiency and accuracy. This method outperforms strong vision backbones across various benchmarks, ...

🔹 Publication Date: Published on Feb 9

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.08683
• PDF: https://arxiv.org/pdf/2602.08683
• Project Page: https://www.lmms-lab.com/onevision-encoder/index.html
• Github: https://github.com/EvolvingLMMs-Lab/OneVision-Encoder/blob/main/docs/data_card.md

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

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

#MultimodalAI #ComputerVision #DeepLearning #Sparsity #AIResearch

✨Intelligent AI Delegation

📝 Summary:
AI agents require better task decomposition and robust delegation. This paper proposes an adaptive framework for intelligent AI delegation, incorporating authority transfer, responsibility, and trust to handle dynamic environments and failures in complex AI and human networks.

🔹 Publication Date: Published on Feb 12

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.11865
• PDF: https://arxiv.org/pdf/2602.11865

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

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

#AIDelegation #AIagents #TaskDecomposition #HumanAICollaboration #MultiAgentSystems

✨ABot-M0: VLA Foundation Model for Robotic Manipulation with Action Manifold Learning

📝 Summary:
ABot-M0 presents a unified framework for embodied agent development that standardizes diverse robotic data and employs action manifold learning to improve prediction efficiency and stability. AI-gener...

🔹 Publication Date: Published on Feb 11

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.11236
• PDF: https://arxiv.org/pdf/2602.11236
• Project Page: https://amap-cvlab.github.io/ABot-Manipulation
• Github: https://github.com/amap-cvlab/ABot-Manipulation

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨SciAgentGym: Benchmarking Multi-Step Scientific Tool-use in LLM Agents

📝 Summary:
SciAgentGym and SciAgentBench enable evaluation of scientific tool-use capabilities, while SciForge improves agent performance through dependency graph modeling of tool interactions. AI-generated summ...

🔹 Publication Date: Published on Feb 13

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.12984
• PDF: https://arxiv.org/pdf/2602.12984
• Github: https://github.com/CMarsRover/SciAgentGYM

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨FLAC: Maximum Entropy RL via Kinetic Energy Regularized Bridge Matching

📝 Summary:
FLAC enables maximum entropy RL for generative policies by regulating stochasticity via kinetic energy. It formulates policy optimization as a Generalized Schrödinger Bridge, avoiding explicit action density estimation while achieving strong performance.

🔹 Publication Date: Published on Feb 13

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.12829
• PDF: https://arxiv.org/pdf/2602.12829
• Project Page: https://pinkmoon-io.github.io/flac.github.io/

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

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

#ReinforcementLearning #MachineLearning #GenerativeAI #OptimalTransport #KineticEnergy

✨Xiaomi-Robotics-0: An Open-Sourced Vision-Language-Action Model with Real-Time Execution

📝 Summary:
Xiaomi-Robotics-0 is an open-sourced vision-language-action model enabling real-time, high-performance robot manipulation. It leverages large-scale pre-training and specialized methods for fast execution on real robots, achieving SOTA simulation and high real-robot success.

🔹 Publication Date: Published on Feb 13

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.12684
• PDF: https://arxiv.org/pdf/2602.12684
• Project Page: https://xiaomi-robotics-0.github.io/
• Github: https://github.com/XiaomiRobotics/Xiaomi-Robotics-0

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

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

#Robotics #AI #VisionLanguageModels #OpenSource #RobotManipulation

✨On Robustness and Chain-of-Thought Consistency of RL-Finetuned VLMs

📝 Summary:
RL-finetuned VLMs are highly vulnerable to misleading text, severely impacting robustness and confidence. RL fine-tuning presents an accuracy-faithfulness trade-off, eroding reasoning reliability despite accuracy gains. This necessitates joint evaluation of correctness, robustness, and reasoning ...

🔹 Publication Date: Published on Feb 13

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.12506
• PDF: https://arxiv.org/pdf/2602.12506

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

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

#VLM #Robustness #ReinforcementLearning #ChainOfThought #AI

✨TADA! Tuning Audio Diffusion Models through Activation Steering

📝 Summary:
Research reveals that specific attention layers in audio diffusion models control distinct musical concepts, enabling precise manipulation of audio features through activation steering. AI-generated s...

🔹 Publication Date: Published on Feb 12

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.11910
• PDF: https://arxiv.org/pdf/2602.11910
• Project Page: https://audio-steering.github.io
• Github: https://github.com/luk-st/steer-audio

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Light4D: Training-Free Extreme Viewpoint 4D Video Relighting

📝 Summary:
Light4D enables consistent 4D video synthesis under target illumination through disentangled flow guidance and temporal consistent attention mechanisms. AI-generated summary Recent advances in diffusi...

🔹 Publication Date: Published on Feb 12

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.11769
• PDF: https://arxiv.org/pdf/2602.11769
• Project Page: https://aigeeksgroup.github.io/Light4D
• Github: https://aigeeksgroup.github.io/Light4D

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Code2Worlds: Empowering Coding LLMs for 4D World Generation

📝 Summary:
Code2Worlds empowers coding LLMs to generate 4D dynamic scenes by formulating it as language-to-simulation code. It uses a dual-stream architecture and physics-aware closed-loop refinement to ensure physical fidelity. The system significantly outperforms baselines, uniquely generating realistic, ...

🔹 Publication Date: Published on Feb 12

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.11757
• PDF: https://arxiv.org/pdf/2602.11757
• Project Page: https://aigeeksgroup.github.io/Code2Worlds
• Github: https://aigeeksgroup.github.io/Code2Worlds

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

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

#LLM #CodeGeneration #4DGeneration #AISimulation #Research

✨GeneralVLA: Generalizable Vision-Language-Action Models with Knowledge-Guided Trajectory Planning

📝 Summary:
GeneralVLA is a hierarchical vision-language-action model that enables zero-shot robotic manipulation through knowledge-guided trajectory planning. It requires no real-world data collection and outperforms existing methods, also generating robust training data.

🔹 Publication Date: Published on Feb 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.04315
• PDF: https://arxiv.org/pdf/2602.04315
• Project Page: https://aigeeksgroup.github.io/GeneralVLA
• Github: https://aigeeksgroup.github.io/GeneralVLA

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Less is Enough: Synthesizing Diverse Data in Feature Space of LLMs

📝 Summary:
Feature Activation Coverage measures data diversity in an interpretable feature space and enables diversity-driven data synthesis that improves downstream performance across multiple language model ar...

🔹 Publication Date: Published on Feb 11

🔹 Paper Links:
• arXiv Page: https://arxiv.org/pdf/2602.10388
• PDF: https://arxiv.org/pdf/2602.10388
• Project Page: https://website-sigma-three-35.vercel.app/
• Github: https://github.com/Zhongzhi660/FAC-Synthesis

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨What does RL improve for Visual Reasoning? A Frankenstein-Style Analysis

📝 Summary:
Reinforcement learning (RL) with verifiable rewards has become a standard post-training stage for boosting visual reasoning in vision-language models, yet it remains unclear what capabilities RL actua...

🔹 Publication Date: Published on Feb 12

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.12395
• PDF: https://arxiv.org/pdf/2602.12395
• Project Page: https://github.com/tianyi-lab/Frankenstein
• Github: https://github.com/tianyi-lab/Frankenstein

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨CoPE-VideoLM: Codec Primitives For Efficient Video Language Models

📝 Summary:
Video Language Models (VideoLMs) empower AI systems to understand temporal dynamics in videos. To fit to the maximum context window constraint, current methods use keyframe sampling which can miss bot...

🔹 Publication Date: Published on Feb 13

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.13191
• PDF: https://arxiv.org/pdf/2602.13191
• Project Page: https://sayands.github.io/cope/
• Github: https://sayands.github.io/cope/

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨BPDQ: Bit-Plane Decomposition Quantization on a Variable Grid for Large Language Models

📝 Summary:
Bit-Plane Decomposition Quantization (BPDQ) improves low-bit quantization by using variable quantization grids derived from bit-planes and scalar coefficients, achieving better accuracy than tradition...

🔹 Publication Date: Published on Feb 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.04163
• PDF: https://arxiv.org/pdf/2602.04163

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨RLinf-Co: Reinforcement Learning-Based Sim-Real Co-Training for VLA Models

📝 Summary:
Reinforcement learning-based sim-real co-training framework improves vision-language-action policy performance through interactive simulation and real-world data anchoring. AI-generated summary Simula...

🔹 Publication Date: Published on Feb 13

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.12628
• PDF: https://arxiv.org/pdf/2602.12628

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨DICE: Diffusion Large Language Models Excel at Generating CUDA Kernels

📝 Summary:
Diffusion large language models (dLLMs) for CUDA kernel generation achieve superior performance through a specialized dataset and reinforcement learning framework. AI-generated summary Diffusion large...

🔹 Publication Date: Published on Feb 12

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.11715
• PDF: https://arxiv.org/pdf/2602.11715
• Project Page: https://deadlykitten4.github.io/DICE/
• Github: https://github.com/deadlykitten4/DICE

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

Limited Time Offer: Premium Q1 & Q2 Publications at Just $300!
🎓 Exclusive February Sale - Ending Soon!
Are you looking to boost your academic profile with high-impact publications? We're offering an exceptional opportunity you don't want to miss!
What We Offer:
✅ Q1 & Q2 Journal Articles - Top-tier, indexed publications
✅ Unbeatable Price: Only $300 per article
✅ Limited Time: Offer valid until the end of February 2026
Why Choose Our Service?

Fast publication process
Reputable Q1 & Q2 journals
Expert support throughout
Guaranteed acceptance

@Omidyzd62

✨Self-EvolveRec: Self-Evolving Recommender Systems with LLM-based Directional Feedback

📝 Summary:
Self-EvolveRec improves recommender system design via a directional feedback loop. It uses a User Simulator for qualitative critiques and a Model Diagnosis Tool for quantitative verification, with adaptive evaluation. It outperforms existing methods.

🔹 Publication Date: Published on Feb 13

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.12612
• PDF: https://arxiv.org/pdf/2602.12612
• Github: https://github.com/Sein-Kim/self_evolverec

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

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

#RecommenderSystems #LLM #MachineLearning #ArtificialIntelligence #DeepLearning

✨AI-Trader: Benchmarking Autonomous Agents in Real-Time Financial Markets

📝 Summary:
AI-Trader introduces the first fully automated live benchmark for evaluating LLM agents in financial decision-making. It reveals that general AI does not ensure trading success, with most agents showing poor returns and weak risk management. Risk control proves crucial, and liquid markets offer b...

🔹 Publication Date: Published on Dec 1, 2025

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2512.10971
• PDF: https://arxiv.org/pdf/2512.10971
• Project Page: https://ai4trade.ai/
• Github: https://github.com/HKUDS/AI-Trader

✨ Datasets citing this paper:
• https://huggingface.co/datasets/T1anyu/AI-Trader

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

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

#AI #LLMAgents #FinTech #AlgorithmicTrading #FinancialAI

✨Quantized Evolution Strategies: High-precision Fine-tuning of Quantized LLMs at Low-precision Cost

📝 Summary:
Quantized LLMs are difficult to fine-tune directly using existing methods. Quantized Evolution Strategies QES enables full-parameter fine-tuning of quantized LLMs. It uses error feedback and seed replay for high-precision optimization at low memory cost, outperforming prior methods.

🔹 Publication Date: Published on Feb 3

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.03120
• PDF: https://arxiv.org/pdf/2602.03120
• Github: https://github.com/dibbla/Quantized-Evolution-Strategies

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

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

#LLM #Quantization #FineTuning #EvolutionStrategies #AI