Any 3D Scene is Worth 1K Tokens: 3D-Grounded Representation for Scene Generation at Scale Paper • 2604.11331 • Published Oct 19 • 5
HSPR: Recovering Software-Agent Benchmark Scores via Hierarchical Posterior Inference Paper • 2610.24611 • Published Oct 17 • 8
Text-to-Image Generative Modeling Without Language Pretraining Paper • 2610.26863 • Published Oct 19 • 8
HERA: Harness-Environment Co-Evolution for Reliable Agentic Abstention Paper • 2610.06563 • Published Oct 5 • 7
Learning to Discriminate Scene Structures Makes Self-Supervised Depth Learning Scalable Paper • 2610.23600 • Published Oct 16 • 9
RSIFlow: Adaptive Self-Improvement Workflows with an Evolving Meta-Agent Paper • 2610.21791 • Published Oct 15 • 9
Progress-Aware Distillation for Mitigating Stagnation in Small Language Model Agents Paper • 2610.19439 • Published Oct 17 • 9
Training Reasoning Models to Generate Diverse Scientific Ideas Paper • 2610.22324 • Published Oct 15 • 8
Predicting Privileged Information Effectiveness for On-Policy Self-Distillation Paper • 2610.22512 • Published Oct 15 • 11
ACEdit: Stable Multi-Turn Image Editing via Asymmetric Conditioning Paper • 2610.22328 • Published Oct 15 • 13
Frechet Forcing: Training Long-Horizon Streaming World Models from Short Clips Paper • 2610.25666 • Published Oct 19 • 14
State World Models: Execute the Rules, Generate the Pixels Paper • 2610.26811 • Published Oct 19 • 15
Interlock: From Local Correctness to Task Completion in Long-Horizon Agents Paper • 2610.19686 • Published Oct 14 • 17