✨QEDBENCH: Quantifying the Alignment Gap in Automated Evaluation of University-Level Mathematical Proofs

📝 Summary:
QEDBench quantifies a systematic Alignment Gap in LLM evaluation of university-level math proofs. It reveals score inflation in some frontier models and a critical reasoning gap in discrete mathematics for others. QEDBench is now a public benchmark.

🔹 Publication Date: Published on Feb 24

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2602.20629
• PDF: https://arxiv.org/pdf/2602.20629
• Project Page: https://quanquancliu.com/QEDBench/index.html
• Github: https://github.com/qqliu/Yale-QEDBench

✨ Datasets citing this paper:
• https://huggingface.co/datasets/qqggez/QEDBench

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨ParEVO: Synthesizing Code for Irregular Data: High-Performance Parallelism through Agentic Evolution

📝 Summary:
ParEVO synthesizes high-performance parallel algorithms for irregular data structures using specialized LLMs and evolutionary coding to achieve significant speedups over existing methods. AI-generated...

🔹 Publication Date: Published on Mar 3

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.02510
• PDF: https://arxiv.org/pdf/2603.02510
• Project Page: https://quanquancliu.com/ParEVO/index.html
• Github: https://github.com/WildAlg/ParEVO

🔹 Models citing this paper:
• https://huggingface.co/qqggez/qwen3-30b-sft-stage2-merged
• https://huggingface.co/qqggez/deepseek-parlay-6.7b

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨LFPO: Likelihood-Free Policy Optimization for Masked Diffusion Models

📝 Summary:
LFPO trains diffusion LLMs by directly optimizing denoising logits through geometric velocity rectification, avoiding complex likelihood calculations. This yields precise gradient estimates, faster inference, and better performance on code and reasoning tasks.

🔹 Publication Date: Published on Mar 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.01563
• PDF: https://arxiv.org/pdf/2603.01563
• Github: https://github.com/kithib/LFPO

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Words & Weights: Streamlining Multi-Turn Interactions via Co-Adaptation

📝 Summary:
ROSA2 framework addresses multi-turn LLM interactions by jointly optimizing textual intent clarity and model parameters, achieving superior mathematical reasoning performance with fewer interaction ro...

🔹 Publication Date: Published on Mar 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.01375
• PDF: https://arxiv.org/pdf/2603.01375
• Github: https://github.com/kithib/ROSA2

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Helios: Real Real-Time Long Video Generation Model

📝 Summary:
Helios is a 14B autoregressive diffusion model that achieves real-time minute-scale video generation at 19.5 FPS on a single GPU. It innovatively overcomes long-video drifting and real-time performance challenges without conventional acceleration or anti-drifting techniques. Helios supports T2V, ...

🔹 Publication Date: Published on Mar 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.04379
• PDF: https://arxiv.org/pdf/2603.04379
• Github: https://pku-yuangroup.github.io/Helios-Page/

🔹 Models citing this paper:
• https://huggingface.co/BestWishYsh/Helios-Base
• https://huggingface.co/BestWishYsh/Helios-Distilled
• https://huggingface.co/BestWishYsh/Helios-Mid

✨ Spaces citing this paper:
• https://huggingface.co/spaces/multimodalart/Helios-Distilled

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨ArtHOI: Articulated Human-Object Interaction Synthesis by 4D Reconstruction from Video Priors

📝 Summary:
ArtHOI synthesizes articulated human-object interactions by formulating 4D reconstruction from monocular video priors, using optical flow for part segmentation and a decoupled reconstruction pipeline ...

🔹 Publication Date: Published on Mar 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.04338
• PDF: https://arxiv.org/pdf/2603.04338
• Project Page: https://arthoi.github.io/
• Github: https://github.com/Inso-13/ArtHOI

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Memex(RL): Scaling Long-Horizon LLM Agents via Indexed Experience Memory

📝 Summary:
A memory mechanism called Memex enables large language model agents to handle long-horizon tasks more effectively by maintaining compact context through structured summaries while storing full interac...

🔹 Publication Date: Published on Mar 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.04257
• PDF: https://arxiv.org/pdf/2603.04257

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Phi-4-reasoning-vision-15B Technical Report

📝 Summary:
A compact open-weight multimodal reasoning model is presented that achieves competitive performance through careful architecture design, high-quality data curation, and a hybrid approach combining dir...

🔹 Publication Date: Published on Mar 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.03975
• PDF: https://arxiv.org/pdf/2603.03975
• Project Page: https://www.microsoft.com/en-us/research/blog/phi-4-reasoning-vision-and-the-lessons-of-training-a-multimodal-reasoning-model/
• Github: https://github.com/microsoft/Phi-4-reasoning-vision-15B

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Heterogeneous Agent Collaborative Reinforcement Learning

📝 Summary:
HACRL enables collaborative reinforcement learning where heterogeneous agents share verified rollouts during training to improve collectively while maintaining independent operation at inference time,...

🔹 Publication Date: Published on Mar 3

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.02604
• PDF: https://arxiv.org/pdf/2603.02604
• Project Page: https://zzx-peter.github.io/hacrl/
• Github: https://zzx-peter.github.io/hacrl/

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨T2S-Bench & Structure-of-Thought: Benchmarking and Prompting Comprehensive Text-to-Structure Reasoning

📝 Summary:
Structure of Thought prompting technique enhances language model performance by guiding explicit intermediate text structuring across diverse tasks, while T2S-Bench benchmark evaluates and improves te...

🔹 Publication Date: Published on Mar 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.03790
• PDF: https://arxiv.org/pdf/2603.03790
• Project Page: https://t2s-bench.github.io/T2S-Bench-Page/
• Github: https://t2s-bench.github.io/T2S-Bench-Page/

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨CubeComposer: Spatio-Temporal Autoregressive 4K 360° Video Generation from Perspective Video

📝 Summary:
CubeComposer is a spatio-temporal autoregressive diffusion model that generates high-resolution 360° panoramic videos by decomposing them into cubemap representations and using efficient autoregressiv...

🔹 Publication Date: Published on Mar 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.04291
• PDF: https://arxiv.org/pdf/2603.04291
• Project Page: https://lg-li.github.io/project/cubecomposer
• Github: https://github.com/TencentARC/CubeComposer

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Proact-VL: A Proactive VideoLLM for Real-Time AI Companions

📝 Summary:
Proact-VL is a multimodal framework that enables real-time interactive AI companions for gaming scenarios with low-latency responses and strong video understanding capabilities. AI-generated summary P...

🔹 Publication Date: Published on Mar 3

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.03447
• PDF: https://arxiv.org/pdf/2603.03447

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨MemSifter: Offloading LLM Memory Retrieval via Outcome-Driven Proxy Reasoning

📝 Summary:
MemSifter is a framework that uses a small proxy model to offload memory retrieval from large language models, employing reinforcement learning with task-performance rewards and training techniques li...

🔹 Publication Date: Published on Mar 3

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.03379
• PDF: https://arxiv.org/pdf/2603.03379
• Github: https://github.com/plageon/MemSifter

🔹 Models citing this paper:
• https://huggingface.co/zstanjj/MemSifter-4B-Thinking

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨MUSE: A Run-Centric Platform for Multimodal Unified Safety Evaluation of Large Language Models

📝 Summary:
MUSE is an open-source platform for evaluating multimodal safety in large language models, incorporating automated cross-modal attack generation and a dual-metric framework to assess alignment across ...

🔹 Publication Date: Published on Mar 3

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.02482
• PDF: https://arxiv.org/pdf/2603.02482

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨SWE-CI: Evaluating Agent Capabilities in Maintaining Codebases via Continuous Integration

📝 Summary:
SWE-CI presents a repository-level benchmark for evaluating code generation agents' ability to maintain code quality through long-term software evolution cycles. AI-generated summary Large language mo...

🔹 Publication Date: Published on Mar 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.03823
• PDF: https://arxiv.org/pdf/2603.03823

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨RIVER: A Real-Time Interaction Benchmark for Video LLMs

📝 Summary:
RIVER Bench is introduced to evaluate real-time video comprehension through retrospective memory, live-perception, and proactive anticipation tasks. This benchmark reveals current offline models struggle with real-time processing, long-term memory, and future perception, highlighting the need for...

🔹 Publication Date: Published on Mar 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.03985
• PDF: https://arxiv.org/pdf/2603.03985
• Github: https://github.com/OpenGVLab/RIVER

✨ Datasets citing this paper:
• https://huggingface.co/datasets/nanamma/RIVER

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨EmbodiedSplat: Online Feed-Forward Semantic 3DGS for Open-Vocabulary 3D Scene Understanding

📝 Summary:
EmbodiedSplat provides real-time 3D scene understanding, combining online 3D Gaussian Splatting with CLIP embeddings from streaming images. It simultaneously reconstructs and semantically comprehends 3D scenes using a novel sparse coefficients field and CLIP global codebook for efficiency and gen...

🔹 Publication Date: Published on Mar 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.04254
• PDF: https://arxiv.org/pdf/2603.04254
• Project Page: https://0nandon.github.io/EmbodiedSplat/
• Github: https://github.com/0nandon/EmbodiedSplat

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

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

#3DSceneUnderstanding #3DGaussianSplatting #ComputerVision #AI #NeuralRendering

Over 20 free courses are now available on our channel for a very limited time.

https://t.iss.one/DataScienceC

✨GroupEnsemble: Efficient Uncertainty Estimation for DETR-based Object Detection

📝 Summary:
DETR models lack spatial uncertainty and current estimation methods are too costly. GroupEnsemble efficiently estimates uncertainty by using independent query groups in a single forward pass with an attention mask. This outperforms Deep Ensembles at a fraction of the cost.

🔹 Publication Date: Published on Mar 2

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.01847
• PDF: https://arxiv.org/pdf/2603.01847

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

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

#ObjectDetection #UncertaintyEstimation #DETR #ComputerVision #MachineLearning

✨InfinityStory: Unlimited Video Generation with World Consistency and Character-Aware Shot Transitions

📝 Summary:
This paper introduces InfinityStory, a novel framework, dataset, and model for long-form video generation. It tackles challenges in background consistency and seamless multi-subject transitions, achieving high consistency and smoother transitions on VBench.

🔹 Publication Date: Published on Mar 4

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2603.03646
• PDF: https://arxiv.org/pdf/2603.03646

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

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

#VideoGeneration #GenerativeAI #DeepLearning #AIResearch #ComputerVision