DC-SAE: Deep Compression Semantic Autoencoder for Faster Diffusion Convergence Paper • 2609.39222 • Published Sep 30 • 24
Post-Training Leaves Behavioral Shadows on Unrelated Decisions Paper • 2609.29233 • Published Sep 24 • 45
Improving Test-Time Scaling with Adaptive Looped Transformers Paper • 2609.35748 • Published Sep 28 • 64
Post-Training Leaves Behavioral Shadows on Unrelated Decisions Paper • 2609.29233 • Published Sep 24 • 45
TimeLens2: Generalist Video Temporal Grounding with Multimodal LLMs Paper • 2607.17423 • Published Jul 19 • 77