Haoxiang Zhang

IPF's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

upvoted 3 papers about 3 hours ago

Rethinking Cross-Tokenizer On-Policy Distillation: From Alignment Coverage to Supervision Reliability

Paper • 2610.08448 • Published 5 days ago • 203

WorldAuditBench: Interactive 3D World Auditing with Multimodal Agents

Paper • 2609.40325 • Published 11 days ago • 106

The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation

Paper • 2609.36484 • Published 12 days ago • 553

upvoted 3 papers about 4 hours ago

TokenRouter: Efficient Serving System for Token-Level LLM Routing

Paper • 2610.12242 • Published 3 days ago • 130

Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight

Paper • 2610.08077 • Published 5 days ago • 156

Latent evolving World Action Model

Paper • 2609.27455 • Published 18 days ago • 2

upvoted a paper about 5 hours ago

AgentGarten: Code Worlds for Evolving Agents

Paper • 2610.12374 • Published 3 days ago • 149

upvoted a paper about 7 hours ago

Raven: The Harness of Harnesses for Composable Agentic Intelligence

Paper • 2609.33439 • Published 14 days ago • 675

upvoted 2 papers 1 day ago

Learning from Teacher Continuations at Student States

Paper • 2609.36246 • Published 13 days ago • 41

Agent Plasticity: Measuring Self-Improvement Through Experience

Paper • 2610.08902 • Published 5 days ago • 5

upvoted 8 papers 2 days ago

Does Learning Protein Folding Generalize to Broader Reasoning?

Paper • 2609.38879 • Published 11 days ago • 62

MiMo-V2.6: Scaling Reinforcement Learning Towards Self-Improvement

Paper • 2610.11959 • Published 3 days ago • 72

CoTrace: Data Recipes for Training Terminal Agents with Harness-Model Co-Evolution

Paper • 2610.10426 • Published 4 days ago • 9

BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agent

Paper • 2508.06600 • Published Aug 8, 2025 • 45

nanoMuse: An Open-Source Personal Agent for Every Device You Own

Paper • 2610.08699 • Published 5 days ago • 101

Prospective Hindsight: Self-Calibrating Reinforcement Learning via Prediction-Reality Gaps

Paper • 2610.02740 • Published 9 days ago • 1

Recursive Game Creator: An Agentic Product-Level Experience-Oriented Game Harness

Paper • 2610.08621 • Published 5 days ago • 87

QiYao-M: Multimodal Time Series Foundation Model with Role-Aware Modeling of Endogenous and Exogenous Modalities

Paper • 2609.34842 • Published 13 days ago • 5

upvoted a paper 3 days ago

STEPQuant: When and Where Errors Matter in Delta-Rule Recurrent State Quantization

Paper • 2609.38169 • Published 12 days ago • 118

upvoted a paper 5 days ago

Sharpening Tax in Post-Training

Paper • 2610.01509 • Published 10 days ago • 111