MaskAlign: Token-Subset Representation Alignment for Efficient Diffusion Training
Paper • 2606.08788 • Published • 4
Urro
urroxyz
AI & ML interests
computational linguistics major 🤖🔎ðŸ”
i am autistic. if i come off rude, i probably didn't mean to. please feel free to ask me for clarification.
Recent Activity
upvoted a paper 3 days ago
MaskAlign: Token-Subset Representation Alignment for Efficient Diffusion TrainingOrganizations
upvoted 8
papers 3 days ago
SG-OPD: Sign-Gated On-Policy Distillation via Sign-Consistency Gating and Phased Teacher Sampling
Paper • 2606.09304 • Published • 6
Risk Under Pressure: Compute-Aware Evaluation of Adversarial Robustness in Language Models
Paper • 2606.11409 • Published • 8
Demystifying Hidden-State Recurrence: Switchable Latent Reasoning with On-Policy Reinforcement Learning
Paper • 2606.13106 • Published • 21
N-GRPO: Embedding-Level Neighbor Mixing for Enhanced Policy Optimization
Paper • 2606.10768 • Published • 24
MaxProof: Scaling Mathematical Proof with Generative-Verifier RL and Population-Level Test-Time Scaling
Paper • 2606.13473 • Published • 84
Robust-U1: Can MLLMs Self-Recover Corrupted Visual Content for Robust Understanding?
Paper • 2606.08063 • Published • 75
MiniMax Sparse Attention
Paper • 2606.13392 • Published • 129
upvoted 5
papers 4 days ago
Reason, Then Re-reason: Cross-view Revisiting Improves Spatial Reasoning
Paper • 2606.11683 • Published • 30
TRL-Bench: Standardizing Cross-Paradigm Representation-Level Evaluation of Tabular Encoders
Paper • 2606.09323 • Published • 50
Beyond Scalar Rewards by Internalizing Reasoning into Score Distributions
Paper • 2606.09076 • Published • 58
Toward Generalist Autonomous Research via Hypothesis-Tree Refinement
Paper • 2606.11926 • Published • 110
Redesign Mixture-of-Experts Routers with Manifold Power Iteration
Paper • 2606.12397 • Published • 85
upvoted 7
papers 6 days ago
Trajectory-Refined Distillation
Paper • 2606.08432 • Published • 7
Why Muon Outperforms Adam: A Curvature Perspective
Paper • 2606.04662 • Published • 10
A Geometric Account of Activation Steering through Angle-Norm Decomposition
Paper • 2606.06735 • Published • 21
Whisper Hallucination Detection and Mitigation via Hidden Representation Steering and Sparse AutoEncoders
Paper • 2606.07473 • Published • 15
End-to-End Context Compression at Scale
Paper • 2606.09659 • Published • 25
Human Psychometric Questionnaires Mischaracterize LLM Behavior
Paper • 2509.10078 • Published • 35
FlashMemory-DeepSeek-V4: Lightning Index Ultra-Long Context via Lookahead Sparse Attention
Paper • 2606.09079 • Published • 61