AI Papers Explained
40
구독자
38
영상
최근 영상
LONGER: Scaling Ultra-Long Sequence Modeling for Industrial Recommenders at ByteDance
AIRA-Compose and AIRA-Design: LLM Agents Discover Novel Neural Architectures Beyond Transformer
LIFE: A Unified Survey of Collaboration, Failure Attribution, and Self-Evolution in LLM Agents
SilverTorch: Unified GPU Model-Based Serving for Large-Scale Recommendation at Meta
Explicit n^1.014 Lower Bound for the Erdős Unit Distance Problem via Golod-Shafarevich
Code as Agent Harness: A Unified View of Executable, Verifiable, Stateful Agent Systems
The RL Conductor: Training a 7B Model to Orchestrate LLM Agents via Reinforcement Learning
HeavySkill: Internalizing Parallel Reasoning and Summarization as an Inner LLM Skill
PLUM: Adapting Pre-trained LLMs for YouTube-Scale Generative Recommendations
OneRec: Unifying Retrieval and Ranking with a Generative Recommender and DPO Alignment
HSTU: Trillion-Parameter Generative Recommenders That Beat DLRMs at Scale
Constitutional AI: Training Harmless Assistants with AI Feedback Instead of Human Labels
PaLM: Scaling a 540B Parameter Language Model with Pathways
Chinchilla: Training Compute-Optimal Large Language Models
Vision Transformer (ViT): Transformers for Image Recognition at Scale
Recurrent Neural Network Regularization: Applying Dropout to LSTMs
Generative Adversarial Nets: Goodfellow et al.'s Original GAN Paper Explained
Latent Diffusion Models: High-Resolution Image Synthesis in Compressed Latent Space
Batch Normalization: Reducing Internal Covariate Shift to Accelerate Deep Network Training
Neural Turing Machines: Differentiable Memory for Learning Algorithms
Playing Atari with Deep Reinforcement Learning: The Original DQN Paper
Auto-Encoding Variational Bayes: The Original VAE Paper by Kingma and Welling
Bahdanau et al. (2014): Neural Machine Translation by Jointly Learning to Align and Translate
Adam: A Method for Stochastic Optimization (Kingma & Ba, 2015)
Scaling Laws for Neural Language Models: Power-Law Trends in Loss, Size, Data, and Compute
LoRA: Low-Rank Adaptation for Efficient Fine-Tuning of Large Language Models
The Coffee Automaton: Quantifying the Rise and Fall of Complexity in Closed Systems
Deep Residual Learning for Image Recognition: The ResNet Paper Explained
Order Matters: Extending Seq2Seq to Handle Sets as Inputs and Outputs
Denoising Diffusion Probabilistic Models (DDPM): High-Quality Image Synthesis Explained