What does it actually take to train a high-performance LLM today?

A Definition of AGI

Scaling Latent Reasoning via Looped Language Models

Large Language Models Report Subjective Experience Under Self-Referential Processing

Nested Learning: The Illusion of Deep Learning Architectures

BADAS: Context Aware Collision Prediction Using Real-World Dashcam Data

Antislop: A Comprehensive Framework for Identifying and Eliminating Repetitive Patterns in LLM

Deep sequence models tend to memorize geometrically; it is unclear why

Radiology's Last Exam (RadLE):Benchmarking Frontier Multimodal AI Against Human Experts in Radiology

LongLive: Real-time Interactive Long Video Generation

KIMI Linear: an expressive, efficient attention architecture

Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models

CMT-Benchmark: A Benchmark for Condensed Matter Theory Built by Expert Researchers

Can Large Language Models Develop Gambling Addiction?

Wan-Alpha: High-Quality Text-to-Video Generation with Alpha Channel

Agent Learning via Early Experience

Moloch's Bargain: Emergent Misalignment When LLMs Compete for Audiences

Pretraining Large Language Models with NVFP4

Less is More: Recursive Reasoning with Tiny Networks

To Infinity and Beyond: Tool-Use Unlocks Length Generalization in State Space Models

The Dragon Hatchling: The Missing Link between the Transformer and Models of the Brai

Mamba-3: Improved Sequence Modeling using State Space Principles

Determination of the fifth Busy Beaver value

Compression( Richard Sutton – Father of RL thinks LLMs are a dead end )

FlowRL: Matching Reward Distributions for LLM Reasoning

Paper2Agent: Reimagining Research Papers As Interactive and Reliable AI Agents

VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Model

DCPO: Dynamic Clipping Policy Optimization

My Favorite Streamer is an LLM: Discovering, Bonding, and Co-Creating in AI VTuber Fandom

DREAMSTATE: Diffusing States and Parameters for Recurrent Large Language Models

Scaling Agents via Continual Pre-training

Adaptive LLM Routing under Budget Constraints

RAPTOR: A Foundation Policy for Quadrotor Control

RL's Razor: Why Online Reinforcement Learning Forgets Less

UI-TARS-2 Technical Report: Advancing GUI Agent with Multi-Turn Reinforcement Learning

Reconstruction Alignment Improves Unified Multimodal Models

Drivel-ology: Challenging LLMs with Interpreting Nonsense with Depth

LIMI: Less is More for Agency

A Survey of Reinforcement Learning for Large Reasoning Models

Reverse-Engineered Reasoning for Open-Ended Generation

The Landscape of Agentic Reinforcement Learning for LLMs: A Survey

Virtual Agent Economies

UMO: Scaling Multi-Identity Consistency for Image Customization via Matching Reward

rStar2-Agent: Agentic Reasoning Technical Report

Video models are zero-shot learners and reasoners

EmbodiedOneVision: Interleaved Vision-Text-Action Pretraining for General Robot Control

SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning

Mathematical research with GPT-5: a Malliavin-Stein experiment

MANZANO: A Simple and Scalable Unified Multimodal Model with a Hybrid Vision Tokenizer

The Illusion of Diminishing Returns: Measuring Long Horizon Execution in LLMs

An AI system to help scientists write expert-level empirical software

Sharing is Caring: Efficient LM Post-Training with Collective RL Experience Sharing

On the Theoretical Limitations of Embedding-Based Retrieval

Why Language Models Hallucinate

Analog in-memory computing attention mechanism for fast and energy-efficient large language models

Latent learning: episodic memory complements parametric learning by enabling reuse of experiences

Linear Memory SE(2) Invariant Attention

DiffLoRA: Differential Low-Rank Adapters for Large Language Models

Associative memory inspires improvements for ICL using a novel attention residual stream

On the Expressiveness of Softmax Attention: A Recurrent Neural Network Perspective