✨The Amazing Agent Race: Strong Tool Users, Weak Navigators

📝 Summary:
The Amazing Agent Race benchmark introduces DAG-based puzzles to evaluate LLM agents' navigation and tool-use capabilities beyond traditional linear benchmarks, revealing that navigation errors domina...

🔹 Publication Date: Published on Apr 17

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.10261
• PDF: https://arxiv.org/pdf/2604.10261
• Project Page: https://minnesotanlp.github.io/the-amazing-agent-race/
• Github: https://github.com/minnesotanlp/the-amazing-agent-race

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Universal statistical signatures of evolution in artificial intelligence architectures

📝 Summary:
The study finds that artificial intelligence architectural evolution follows the same statistical patterns as biological evolution, including similar fitness effect distributions and convergence dynam...

🔹 Publication Date: Published on Apr 12

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.10571
• PDF: https://arxiv.org/pdf/2604.10571
• Github: https://github.com/mool32/ai-evolution-universal-signatures

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Motif-Video 2B: Technical Report

📝 Summary:
Motif-Video 2B achieves high text-to-video quality with a specialized architecture and efficient training methods. It uses shared cross-attention and a three-part backbone to outperform larger models using significantly fewer parameters and less data.

🔹 Publication Date: Published on Apr 14

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.16503
• PDF: https://arxiv.org/pdf/2604.16503
• Project Page: https://motiftech.io/videoshowcase

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Stratagem: Learning Transferable Reasoning via Trajectory-Modulated Game Self-Play

📝 Summary:
STRATAGEM addresses limitations in reasoning transfer for language models by using a reasoning transferability coefficient and evolution reward to promote abstract, domain-agnostic patterns over game-...

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.17696
• PDF: https://arxiv.org/pdf/2604.17696
• Github: https://github.com/ydyyyy/Stratagem

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨The Geometric Canary: Predicting Steerability and Detecting Drift via Representational Stability

📝 Summary:
Geometric stability measures predict language model controllability and detect structural degradation, with supervised variants excelling at steering prediction and unsupervised variants at drift dete...

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.17698
• PDF: https://arxiv.org/pdf/2604.17698
• Github: https://github.com/prashantcraju/geometric-canary

🔹 Models citing this paper:
• https://huggingface.co/pcr2120/shesha-geometry

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Geometric coherence of single-cell CRISPR perturbations reveals regulatory architecture and predicts cellular stress

📝 Summary:
G e n o m e e n g i n e e r i n g h a s a c h i e v e d r e m a r k a b l e s e q u e n c e - l e v e l p r e c i s i o n , y e t p r e d i c t i n g t h e t r a n s c r i p t o m i c s t a t e t h a ...

🔹 Publication Date: Published on Apr 17

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.16642
• PDF: https://arxiv.org/pdf/2604.16642
• Github: https://github.com/prashantcraju/geometric-stability-crispr

🔹 Models citing this paper:
• https://huggingface.co/pcr2120/shesha-geometry

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Revisiting a Pain in the Neck: A Semantic Reasoning Benchmark for Language Models

📝 Summary:
SemanticQA is a new benchmark to evaluate language models on semantic phrase processing, covering various phrase types. It reveals significant performance differences, especially in semantic reasoning tasks, highlighting variations in models comprehension.

🔹 Publication Date: Published on Apr 17

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.16593
• PDF: https://arxiv.org/pdf/2604.16593
• Github: https://github.com/jacklanda/SemanticQA

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Crowded in B-Space: Calibrating Shared Directions for LoRA Merging

📝 Summary:
LoRA adapter merging performance can be improved by separately calibrating the output-side matrix B to reduce interference from shared directions while preserving task-specific information. AI-generat...

🔹 Publication Date: Published on Apr 18

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.16826
• PDF: https://arxiv.org/pdf/2604.16826

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Meta-learning In-Context Enables Training-Free Cross Subject Brain Decoding

📝 Summary:
A meta-optimized approach enables generalizable semantic visual decoding from fMRI by rapidly inferring unique neural encoding patterns from few image-brain examples without fine-tuning across subject...

🔹 Publication Date: Published on Apr 9

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.08537
• PDF: https://arxiv.org/pdf/2604.08537
• Github: https://github.com/ezacngm/brainCodec

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Multiplication in Multimodal LLMs: Computation with Text, Image, and Audio Inputs

📝 Summary:
Multimodal large language models demonstrate consistent computational limitations in exact multi-digit multiplication across different representations and modalities, with performance closely tied to ...

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.18203
• PDF: https://arxiv.org/pdf/2604.18203
• Project Page: https://neuristemic.ai/multiplication-in-multimodal-llms/

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨OneVL: One-Step Latent Reasoning and Planning with Vision-Language Explanation

📝 Summary:
OneVL is a unified vision-language-action framework that improves latent chain-of-thought reasoning for autonomous driving. It uses dual language and visual world model supervision to force latent tokens to internalize causal dynamics, achieving state-of-the-art accuracy at answer-only latency.

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.18486
• PDF: https://arxiv.org/pdf/2604.18486
• Project Page: https://xiaomi-embodied-intelligence.github.io/OneVL/

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Agent-World: Scaling Real-World Environment Synthesis for Evolving General Agent Intelligence

📝 Summary:
Agent-World introduces a self-evolving training framework that advances general agent intelligence through autonomous environment discovery and continuous learning across diverse real-world scenarios....

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.18292
• PDF: https://arxiv.org/pdf/2604.18292
• Project Page: https://agent-tars-world.github.io/-/
• Github: https://agent-tars-world.github.io/-/

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨MultiWorld: Scalable Multi-Agent Multi-View Video World Models

📝 Summary:
MultiWorld is a unified framework for multi-agent multi-view world modeling that achieves accurate multi-agent control while maintaining multi-view consistency through specialized modules for conditio...

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.18564
• PDF: https://arxiv.org/pdf/2604.18564
• Project Page: https://multi-world.github.io/
• Github: https://github.com/CIntellifusion/MultiWorld

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨WebCompass: Towards Multimodal Web Coding Evaluation for Code Language Models

📝 Summary:
WebCompass evaluates web development capabilities through diverse input modalities and task types, using automated evaluation methods that simulate real-world coding workflows. AI-generated summary La...

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.18224
• PDF: https://arxiv.org/pdf/2604.18224

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Precise Debugging Benchmark: Is Your Model Debugging or Regenerating?

📝 Summary:
Frontier LLMs demonstrate high test pass rates but poor precision in debugging tasks, indicating a gap between functional correctness and precise fault localization. AI-generated summary Unlike code c...

🔹 Publication Date: Published on Apr 19

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.17338
• PDF: https://arxiv.org/pdf/2604.17338
• Project Page: https://precise-debugging-benchmark.github.io/
• Github: https://github.com/Bill1235813/PDB

✨ Datasets citing this paper:
• https://huggingface.co/datasets/Precise-Debugging-Benchmarking/PDB-Multi
• https://huggingface.co/datasets/Precise-Debugging-Benchmarking/PDB-Single-Hard
• https://huggingface.co/datasets/Precise-Debugging-Benchmarking/PDB-Single

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨MedConclusion: A Benchmark for Biomedical Conclusion Generation from Structured Abstracts

📝 Summary:
A large-scale dataset of 5.7 million PubMed structured abstracts is introduced for biomedical conclusion generation, enabling evaluation of large language models' ability to reason from structured sci...

🔹 Publication Date: Published on Apr 7

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.06505
• PDF: https://arxiv.org/pdf/2604.06505
• Github: https://github.com/Harvard-AI-and-Robotics-Lab/MedConclusion

✨ Datasets citing this paper:
• https://huggingface.co/datasets/harvardairobotics/MedConclusion-Compact
• https://huggingface.co/datasets/harvardairobotics/MedConclusion

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨MNAFT: modality neuron-aware fine-tuning of multimodal large language models for image translation

📝 Summary:
Modality neuron-aware fine-tuning (MNAFT) enhances image translation by selectively updating specific neurons in multimodal large language models, preserving pre-trained knowledge while improving cros...

🔹 Publication Date: Published on Apr 18

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.16943
• PDF: https://arxiv.org/pdf/2604.16943

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Training LLM Agents for Spontaneous, Reward-Free Self-Evolution via World Knowledge Exploration

📝 Summary:
Agents equipped with intrinsic meta-evolution capabilities demonstrate improved performance on web navigation tasks through self-generated world knowledge without external supervision. AI-generated su...

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.18131
• PDF: https://arxiv.org/pdf/2604.18131
• Github: https://github.com/Bklight999/world-knowledge

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨ClawEnvKit: Automatic Environment Generation for Claw-Like Agents

📝 Summary:
An automated pipeline generates diverse, verified environments for claw-like agents from natural language descriptions, enabling large-scale benchmark construction and continuous evaluation. AI-genera...

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.18543
• PDF: https://arxiv.org/pdf/2604.18543
• Github: https://github.com/xirui-li/ClawEnvKit

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨MathNet: a Global Multimodal Benchmark for Mathematical Reasoning and Retrieval

📝 Summary:
MathNet is a large-scale, multilingual, multimodal dataset of Olympiad-level math problems designed for evaluating mathematical reasoning and retrieval in generative models and embedding-based systems...

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.18584
• PDF: https://arxiv.org/pdf/2604.18584
• Project Page: https://mathnet.mit.edu/

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research

✨Modeling Multiple Support Strategies within a Single Turn for Emotional Support Conversations

📝 Summary:
Multi-strategy utterance generation methods for emotional support conversations outperform single-strategy approaches by enabling multiple support strategies within individual utterances. AI-generated...

🔹 Publication Date: Published on Apr 20

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2604.17972
• PDF: https://arxiv.org/pdf/2604.17972
• Project Page: https://github.com/aliyun/qwen-dianjin

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

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

#AI #DataScience #MachineLearning #HuggingFace #Research