arXiv
DOPD: Dual On-policy Distillation
On-policy distillation (OPD) offers superior capacity transfer by supervising student-sampled trajectories with dense...
🤖 AI 前沿资讯
每天自动聚合 AI 领域最新动态
arXiv
Pessimism's Paradox: Conservative Offline Training Amplifies Reward Hacking During Online Adaptation in Reasoning Models
Conservative offline training is widely advocated as a safe foundation for subsequent online adaptation: if a policy ...
arXiv
GROW$^2$: Grounding Which and Where for Robot Tool Use
Can the robot use a plate to cut a cake if no knife is available? Tool use greatly expands robot capabilities, but to...
arXiv
Self-Evolving World Models for LLM Agent Planning
World models offer a principled way to equip long-horizon LLM agents with foresight: predictions of action consequenc...
arXiv
LeVo 2: Stable and Melodious Song Generation via Hierarchical Representation Modeling and Progressive Post-Training
Full-length song generation must preserve coherence and musicality, render detailed vocal and accompaniment acoustics...
arXiv
VLK: Learning Humanoid Loco-Manipulation from Synthetic Interactions in Reconstructed Scenes
Perception-based humanoid loco-manipulation requires connecting egocentric observations and task instructions to whol...
HuggingFace
PoseShield: Neural Collision Fields for Human Self-Collision Resolution
Self-collision remains a persistent challenge in SMPL-based human pose estimation and motion generation. Under extrem...
HuggingFace
One Forward Beats Two: InnerZoom for Accurate and Efficient GUI Grounding
MLLM-based GUI grounding methods commonly formulate target localization as autoregressive coordinate generation, enab...
HuggingFace
SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing
In real-world applications, guardrails are often expected to identify unsafe user-model interactions according to app...
HuggingFace
Walking in the Implicit: Interactive World Exploration via Neural Scene Representation
Interactive video generation systems for camera-controlled world exploration roll out growing sequences of latent vid...
HuggingFace
OSWorld2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks
Existing computer-use benchmarks fail to capture the realism, complexity, and long-horizon demands of real-world comp...
HuggingFace
TUA-Bench: A Benchmark for General-Purpose Terminal-Use Agents
As large language models and harness frameworks continue to advance, agents operating in terminals are increasingly c...