Noticias
Parallax: A Parameterized Local Linear Attention That Keeps Softmax and Adds a Learned Covariance Correction Branch
The Transformer’s attention mechanism has barely changed since 2017. Most efficiency work has tried to...
Paper2Agent: Reimagining Research Papers As Interactive and Reliable AI Agents
arXiv:2509.06917v2 Announce Type: replace-cross Abstract: We introduce Paper2Agent, an automated framework that converts research papers...
PAG: Multi-Turn Reinforced LLM Self-Correction with Policy as Generative Verifier
arXiv:2506.10406v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated impressive capabilities in complex...
PadChest-GR: A Bilingual Chest X-ray Dataset for Grounded Radiology Report Generation
arXiv:2411.05085v2 Announce Type: replace-cross Abstract: Radiology report generation (RRG) aims to create free-text radiology reports...
P-React: Synthesizing Topic-Adaptive Reactions of Personality Traits via Mixture of Specialized LoRA Experts
arXiv:2406.12548v3 Announce Type: replace Abstract: Personalized large language models (LLMs) have attracted great attention in...
OutSafe-Bench: A Benchmark for Multimodal Offensive Content Detection in Large Language Models
arXiv:2511.10287v1 Announce Type: cross Abstract: Since Multimodal Large Language Models (MLLMs) are increasingly being integrated...
ORBIT: Scalable and Verifiable Data Generation for Search Agents on a Tight Budget
arXiv:2604.01195v2 Announce Type: replace Abstract: Search agents, which integrate language models (LMs) with web search...
Optimizing Length Compression in Large Reasoning Models
arXiv:2506.14755v2 Announce Type: replace-cross Abstract: Large Reasoning Models (LRMs) have achieved remarkable success, yet they...
Optimizing Assembly Code with LLMs: Reinforcement Learning Outperforms Traditional Compilers
LLMs have shown impressive capabilities across various programming tasks, yet their potential for program optimization...
Operationalizing AI for Scale and Sovereignty
Companies are taking control of their own data to tailor AI for their needs. The...
OpenVLThinker: Complex Vision-Language Reasoning via Iterative SFT-RL Cycles
arXiv:2503.17352v3 Announce Type: replace-cross Abstract: We introduce OpenVLThinker, one of the first open-source large vision-language...
OpenThoughts: A Scalable Supervised Fine-Tuning SFT Data Curation Pipeline for Reasoning Models
The Growing Complexity of Reasoning Data Curation Recent reasoning models, such as DeepSeek-R1 and o3...
