✨Easy Dataset: A Unified and Extensible Framework for Synthesizing LLM Fine-Tuning Data from Unstructured Documents

📝 Summary:
Easy Dataset is a framework that synthesizes LLM fine-tuning data from unstructured documents using a GUI and LLMs. It generates domain-specific question-answer pairs with human oversight. This improves LLM performance in specific domains while retaining general knowledge.

🔹 Publication Date: Published on Jul 5

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2507.04009
• PDF: https://arxiv.org/pdf/2507.04009
• Github: https://github.com/ConardLi/easy-dataset

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

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

#LLM #DataSynthesis #FineTuning #AI #NLP

✨InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models

📝 Summary:
InternVL3 introduces a native multimodal pre-training paradigm, jointly learning from multimodal and text data to overcome conventional alignment challenges. This unified approach, combined with advanced techniques, achieves state-of-the-art performance on multimodal tasks, rivaling proprietary m...

🔹 Publication Date: Published on Apr 14

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2504.10479
• PDF: https://arxiv.org/pdf/2504.10479
• Project Page: https://internvl.github.io/blog/2025-04-11-InternVL-3.0/

🔹 Models citing this paper:
• https://huggingface.co/OpenGVLab/InternVL3-78B
• https://huggingface.co/OpenGVLab/InternVL3_5-241B-A28B
• https://huggingface.co/OpenGVLab/InternVL3-8B

✨ Datasets citing this paper:
• https://huggingface.co/datasets/OpenGVLab/MMPR-v1.2-prompts

✨ Spaces citing this paper:
• https://huggingface.co/spaces/AntResearchNLP/ViLaBench
• https://huggingface.co/spaces/TIGER-Lab/MEGA-Bench
• https://huggingface.co/spaces/prithivMLmods/Tiny-VLMs-Lab

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

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

#MultimodalAI #DeepLearning #AIResearch #OpenSourceAI #GenerativeAI

✨A decoder-only foundation model for time-series forecasting

📝 Summary:
This paper introduces a decoder-only foundation model, adapted from large language models, for time-series forecasting. It achieves near-optimal zero-shot performance on diverse datasets across various time scales and granularities.

🔹 Publication Date: Published on Oct 14, 2023

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2310.10688
• PDF: https://arxiv.org/pdf/2310.10688
• Github: https://github.com/google-research/timesfm

🔹 Models citing this paper:
• https://huggingface.co/google/timesfm-1.0-200m
• https://huggingface.co/google/timesfm-2.0-500m-pytorch
• https://huggingface.co/google/timesfm-2.5-200m-pytorch

✨ Spaces citing this paper:
• https://huggingface.co/spaces/autogluon/fev-leaderboard
• https://huggingface.co/spaces/JayLacoma/Trader_Technical_Indicators
• https://huggingface.co/spaces/pavel321/huggingface-cli-completion

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

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

#TimeSeriesForecasting #FoundationModels #MachineLearning #DeepLearning #AI

✨RLinf-VLA: A Unified and Efficient Framework for VLA+RL Training

📝 Summary:
RLinf-VLA is a unified framework for scalable reinforcement learning training of vision-language-action models, overcoming supervised fine-tuning limitations. It offers a 1.6x-1.8x speedup, supports diverse architectures and algorithms, and shows strong generalization in simulation and on a real ...

🔹 Publication Date: Published on Oct 8

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.06710
• PDF: https://arxiv.org/pdf/2510.06710
• Project Page: https://rlinf.readthedocs.io/en/latest/
• Github: https://github.com/RLinf/RLinf

🔹 Models citing this paper:
• https://huggingface.co/RLinf/RLinf-math-7B

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

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

#ReinforcementLearning #VLA #Robotics #AIResearch #MachineLearning

✨ChronoEdit: Towards Temporal Reasoning for Image Editing and World Simulation

📝 Summary:
ChronoEdit ensures physical consistency in image editing by reframing it as a video generation problem. It uses pretrained video models and temporal reasoning tokens to imagine plausible physical transformations between edited images. This approach significantly improves realism and visual fideli...

🔹 Publication Date: Published on Oct 5

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.04290
• PDF: https://arxiv.org/pdf/2510.04290
• Project Page: https://research.nvidia.com/labs/toronto-ai/chronoedit
• Github: https://github.com/nv-tlabs/ChronoEdit

🔹 Models citing this paper:
• https://huggingface.co/nvidia/ChronoEdit-14B-Diffusers
• https://huggingface.co/vantagewithai/ChronoEdit-GGUF
• https://huggingface.co/vantagewithai/ChronoEdit-fp8-scaled

✨ Spaces citing this paper:
• https://huggingface.co/spaces/nvidia/ChronoEdit
• https://huggingface.co/spaces/JarlJarle/nvidia-ChronoEdit-14B-Diffusers

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

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

#ImageEditing #VideoGeneration #TemporalReasoning #ComputerVision #AIResearch

✨Less is More: Recursive Reasoning with Tiny Networks

📝 Summary:
Tiny Recursive Model TRM uses a simple, two-layer network for recursive reasoning. It significantly outperforms larger language models on complex puzzle tasks like ARC-AGI, achieving high generalization with vastly fewer parameters. TRM demonstrates superior performance with minimal resources.

🔹 Publication Date: Published on Oct 6

🔹 Paper Links:
• arXiv Page: https://arxivexplained.com/papers/less-is-more-recursive-reasoning-with-tiny-networks
• PDF: https://arxiv.org/pdf/2510.04871
• Project Page: https://alexiajm.github.io/2025/09/29/tiny_recursive_models.html
• Github: https://github.com/SamsungSAILMontreal/TinyRecursiveModels/issues/2

🔹 Models citing this paper:
• https://huggingface.co/wtfmahe/Samsung-TRM
• https://huggingface.co/ordlibrary/X402

✨ Datasets citing this paper:
• https://huggingface.co/datasets/emiliocantuc/sudoku-extreme-1k-aug-1000

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

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

#RecursiveReasoning #TinyAI #EfficientAI #AIResearch #MachineLearning

✨Brain-IT: Image Reconstruction from fMRI via Brain-Interaction Transformer

📝 Summary:
Brain-IT reconstructs high-fidelity images from fMRI using a Brain Interaction Transformer. It surpasses current methods visually and objectively, and requires significantly less training data for new subjects.

🔹 Publication Date: Published on Oct 29

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.25976
• PDF: https://arxiv.org/pdf/2510.25976
• Project Page: https://amitzalcher.github.io/Brain-IT/
• Github: https://amitzalcher.github.io/Brain-IT/

✨ Datasets citing this paper:
• https://huggingface.co/datasets/Amitz244/Brain-IT_Results

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

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

#fMRI #ImageReconstruction #DeepLearning #Neuroscience #BrainIT

✨Can Visual Input Be Compressed? A Visual Token Compression Benchmark for Large Multimodal Models

📝 Summary:
UniPruneBench is a new benchmark for evaluating visual token pruning in large multimodal models LMMs. It standardizes evaluation across tasks and models, revealing that random pruning is a strong baseline and OCR is most sensitive to pruning. The pruning ratio greatly impacts performance.

🔹 Publication Date: Published on Nov 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2511.02650
• PDF: https://arxiv.org/pdf/2511.02650
• Project Page: https://uniprunebench-lmm.github.io/
• Github: https://github.com/TianfanPeng/VLMUniPruneBench

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

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

#LMMs #VisualCompression #DeepLearning #ComputerVision #AIResearch

✨CodeClash: Benchmarking Goal-Oriented Software Engineering

📝 Summary:
CodeClash is a benchmark evaluating language models on open-ended, goal-oriented code development through competitive tournaments. It shows LMs struggle with strategic reasoning and long-term codebase maintenance, performing poorly against human experts.

🔹 Publication Date: Published on Nov 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2511.00839
• PDF: https://arxiv.org/pdf/2511.00839

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

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

#LanguageModels #SoftwareEngineering #AIEvaluation #CodeDevelopment #Benchmarking

✨TabDSR: Decompose, Sanitize, and Reason for Complex Numerical Reasoning in Tabular Data

📝 Summary:
TabDSR improves LLM performance on complex tabular numerical reasoning by decomposing queries, sanitizing tables, and using program-of-thoughts reasoning. It achieves state-of-the-art accuracy, consistently outperforming existing methods.

🔹 Publication Date: Published on Nov 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2511.02219
• PDF: https://arxiv.org/pdf/2511.02219

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

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

#LLM #TabularData #NumericalReasoning #DataScience #AI

✨Don't Blind Your VLA: Aligning Visual Representations for OOD Generalization

📝 Summary:
Naive action fine-tuning degrades visual representations in Vision-Language-Action models. This study analyzes this degradation and introduces a simple method to align representations, improving out-of-distribution generalization.

🔹 Publication Date: Published on Oct 29

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.25616
• PDF: https://arxiv.org/pdf/2510.25616
• Project Page: https://blind-vla-paper.github.io
• Github: https://github.com/CognitiveAISystems/BlindVLA

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

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

#VLA #OODGeneralization #ComputerVision #MachineLearning #RepresentationLearning

✨The Collaboration Gap

📝 Summary:
A new benchmark reveals a collaboration gap where AI models performing well solo degrade significantly when paired. Starting with a stronger agent relay inference helps bridge this gap. This suggests a need for collaboration-aware evaluation and training.

🔹 Publication Date: Published on Nov 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2511.02687
• PDF: https://arxiv.org/pdf/2511.02687

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

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

#AI #Collaboration #MultiAgentSystems #AIResearch #AIEvaluation

✨LightRAG: Simple and Fast Retrieval-Augmented Generation

📝 Summary:
LightRAG improves Retrieval-Augmented Generation by addressing limitations of flat data representations and inadequate contextual awareness. It integrates graph structures into text indexing and retrieval, enhancing accuracy, efficiency, and response times through a dual-level system.

🔹 Publication Date: Published on Oct 8, 2024

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2410.05779
• PDF: https://arxiv.org/pdf/2410.05779
• Github: https://github.com/hkuds/lightrag

✨ Spaces citing this paper:
• https://huggingface.co/spaces/rm-lht/lightrag

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

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

#RAG #AI #NLP #GraphAI #InformationRetrieval

✨RiddleBench: A New Generative Reasoning Benchmark for LLMs

📝 Summary:
RiddleBench, a new benchmark of 1,737 puzzles, reveals fundamental weaknesses in state-of-the-art LLMs, including hallucination cascades and poor self-correction. Models achieve only about 60% accuracy, underscoring the need for more robust and reliable reasoning capabilities.

🔹 Publication Date: Published on Oct 28

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.24932
• PDF: https://arxiv.org/pdf/2510.24932

✨ Datasets citing this paper:
• https://huggingface.co/datasets/ai4bharat/RiddleBench

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

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

#LLMs #GenerativeAI #AIResearch #Benchmarks #NLP

✨AgentScope 1.0: A Developer-Centric Framework for Building Agentic Applications

📝 Summary:
AgentScope 1.0 is a developer-centric framework for building agentic applications. It offers flexible tool-based interactions, unified interfaces, and ReAct-based infrastructure to enable efficient and safe development and deployment.

🔹 Publication Date: Published on Aug 22

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2508.16279
• PDF: https://arxiv.org/pdf/2508.16279
• Github: https://github.com/agentscope-ai/agentscope

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

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

#AIAgents #AIdevelopment #SoftwareFramework #AItools #ReActAI

✨RoboChallenge: Large-scale Real-robot Evaluation of Embodied Policies

📝 Summary:
RoboChallenge is an online evaluation system for robotic control algorithms, especially VLA models. It enables large-scale, reproducible real-robot testing to survey state-of-the-art models.

🔹 Publication Date: Published on Oct 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2510.17950
• PDF: https://arxiv.org/pdf/2510.17950

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

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

#Robotics #AI #MachineLearning #EmbodiedAI #RoboticsEvaluation

✨Reg-DPO: SFT-Regularized Direct Preference Optimization with GT-Pair for Improving Video Generation

📝 Summary:
This paper presents GT-Pair for automatic preference data construction and Reg-DPO, which adds SFT loss to DPO for stable training. Combined with memory optimizations, it significantly improves video generation quality, outperforming existing methods.

🔹 Publication Date: Published on Nov 3

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2511.01450
• PDF: https://arxiv.org/pdf/2511.01450
• Github: https://github.com/JieDuTQS/Reg-DPO

🔹 Models citing this paper:
• https://huggingface.co/dujielvtqs/Reg-DPO

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

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

#VideoGeneration #GenerativeAI #DeepLearning #DPO #AIResearch