Ежедневные Исследовательские Статьи по ИИ

Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAs

Visual-RFT: Visual Reinforcement Fine-Tuning

Difix3D+: Improving 3D Reconstructions with Single-Step Diffusion Models

Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRs

DiffRhythm: Blazingly Fast and Embarrassingly Simple End-to-End Full-Length Song Generation with Latent Diffusion

From Hours to Minutes: Lossless Acceleration of Ultra Long Sequence Generation up to 100K Tokens

OneRec: Unifying Retrieve and Rank with Generative Recommender and Iterative Preference Alignment

When an LLM is apprehensive about its answers -- and when its uncertainty is justified

Liger: Linearizing Large Language Models to Gated Recurrent Structures

Efficient Test-Time Scaling via Self-Calibration

Speculative Ad-hoc Querying

Qilin: A Multimodal Information Retrieval Dataset with APP-level User Sessions

Large-Scale Data Selection for Instruction Tuning

Kiss3DGen: Repurposing Image Diffusion Models for 3D Asset Generation

DuoDecoding: Hardware-aware Heterogeneous Speculative Decoding with Dynamic Multi-Sequence Drafting

SampleMix: A Sample-wise Pre-training Data Mixing Strategey by Coordinating Data Quality and Diversity

CodeArena: A Collective Evaluation Platform for LLM Code Generation

VideoUFO: A Million-Scale User-Focused Dataset for Text-to-Video Generation

PodAgent: A Comprehensive Framework for Podcast Generation

Word Form Matters: LLMs' Semantic Reconstruction under Typoglycemia

AI-Invented Tonal Languages: Preventing a Machine Lingua Franca Beyond Human Understanding

General Reasoning Requires Learning to Reason from the Get-go

Teaching Metric Distance to Autoregressive Multimodal Foundational Models

CLEA: Closed-Loop Embodied Agent for Enhancing Task Execution in Dynamic Environments

RSQ: Learning from Important Tokens Leads to Better Quantized LLMs

Why Are Web AI Agents More Vulnerable Than Standalone LLMs? A Security Analysis

Direct Discriminative Optimization: Your Likelihood-Based Visual Generative Model is Secretly a GAN Discriminator

Unposed Sparse Views Room Layout Reconstruction in the Age of Pretrain Model