Accelerating RL Post-Training Rollouts via System-Integrated Speculative Decoding
Paper • 2604.26779 • Published • 3 •
1
Avi
avahal
AI & ML interests
LLMs
Recent Activity
commentedon a paper about 10 hours ago
Accelerating RL Post-Training Rollouts via System-Integrated Speculative Decoding commentedon a paper about 10 hours ago
Diffusion Templates: A Unified Plugin Framework for Controllable Diffusion commentedon a paper about 10 hours ago
ClawGym: A Scalable Framework for Building Effective Claw AgentsOrganizations
None yet
commented 6
papers about 10 hours ago
Diffusion Templates: A Unified Plugin Framework for Controllable Diffusion
Paper • 2604.24351 • Published • 6 •
2
ClawGym: A Scalable Framework for Building Effective Claw Agents
Paper • 2604.26904 • Published • 37 •
2
Turning the TIDE: Cross-Architecture Distillation for Diffusion Large Language Models
Paper • 2604.26951 • Published • 36 •
2
Large Language Models Explore by Latent Distilling
Paper • 2604.24927 • Published • 53 •
4
GLM-5V-Turbo: Toward a Native Foundation Model for Multimodal Agents
Paper • 2604.26752 • Published • 71 •
2
commented 2
papers about 21 hours ago
commented 7
papers 3 days ago
Contexts are Never Long Enough: Structured Reasoning for Scalable Question Answering over Long Document Sets
Paper • 2604.22294 • Published • 15 •
3
FlowAnchor: Stabilizing the Editing Signal for Inversion-Free Video Editing
Paper • 2604.22586 • Published • 15 •
4
LLM Safety From Within: Detecting Harmful Content with Internal Representations
Paper • 2604.18519 • Published • 21 •
3
DiffNR: Diffusion-Enhanced Neural Representation Optimization for Sparse-View 3D Tomographic Reconstruction
Paper • 2604.21518 • Published • 27 •
3
Video Analysis and Generation via a Semantic Progress Function
Paper • 2604.22554 • Published • 59 •
4
Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond
Paper • 2604.22748 • Published • 213 •
5
From Skills to Talent: Organising Heterogeneous Agents as a Real-World Company
Paper • 2604.22446 • Published • 112 •
5
commented 5
papers 7 days ago
TingIS: Real-time Risk Event Discovery from Noisy Customer Incidents at Enterprise Scale
Paper • 2604.21889 • Published • 12 •
3
Seeing Fast and Slow: Learning the Flow of Time in Videos
Paper • 2604.21931 • Published • 19 •
3
StyleID: A Perception-Aware Dataset and Metric for Stylization-Agnostic Facial Identity Recognition
Paper • 2604.21689 • Published • 24 •
4
UniT: Toward a Unified Physical Language for Human-to-Humanoid Policy Learning and World Modeling
Paper • 2604.19734 • Published • 29 •
3
WorldMark: A Unified Benchmark Suite for Interactive Video World Models
Paper • 2604.21686 • Published • 36 •
3