AI研究論文每日精選

WALL-E: World Alignment by Rule Learning Improves World Model-based LLM Agents

MathCoder2: Better Math Reasoning from Continued Pretraining on Model-translated Mathematical Code

MLLM as Retriever: Interactively Learning Multimodal Retrieval for Embodied Agents

PrefixQuant: Static Quantization Beats Dynamic through Prefixed Outliers in LLMs

Benchmarking Agentic Workflow Generation

DART: Denoising Autoregressive Transformer for Scalable Text-to-Image Generation

Agent S: An Open Agentic Framework that Uses Computers Like a Human

DICE: Discrete Inversion Enabling Controllable Editing for Multinomial Diffusion and Masked Generative Models

Rectified Diffusion: Straightness Is Not Your Need in Rectified Flow

Progressive Autoregressive Video Diffusion Models

GLOV: Guided Large Language Models as Implicit Optimizers for Vision Language Models

Intriguing Properties of Large Language and Vision Models

Everything Everywhere All at Once: LLMs can In-Context Learn Multiple Tasks in Superposition

Preserving Multi-Modal Capabilities of Pre-trained VLMs for Improving Vision-Linguistic Compositionality

Towards Self-Improvement of LLMs via MCTS: Leveraging Stepwise Knowledge with Curriculum Preference Learning

Optima: Optimizing Effectiveness and Efficiency for LLM-Based Multi-Agent System

Scaling Up Your Kernels: Large Kernel Design in ConvNets towards Universal Representations

Emergent properties with repeated examples

SFTMix: Elevating Language Model Instruction Tuning with Mixup Recipe

Cheating Automatic LLM Benchmarks: Null Models Achieve High Win Rates

Accelerated Preference Optimization for Large Language Model Alignment

MotionGS: Exploring Explicit Motion Guidance for Deformable 3D Gaussian Splatting

Vector-ICL: In-context Learning with Continuous Vector Representations

Data Advisor: Dynamic Data Curation for Safety Alignment of Large Language Models

LPZero: Language Model Zero-cost Proxy Search from Zero

Zebra: In-Context and Generative Pretraining for Solving Parametric PDEs