Tacit-TTS: From Autoregressive Decoding to Masked Prediction for Efficient Transcript-Free Voice Cloning
Paper • 2609.38658 • Published • 12
🔄 In a Training Loop
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
View all activity
Organizations
upvoted 8
papers 3 days ago
Working Around the Compute Ceiling: Byte-Exact Memory in Galahad Makes LLM Reading a One-Time Cost LLM Reading a One-Time Cost
Paper • 2609.39358 • Published • 3
BiasReducer: Adaptive Bias Mitigation for Reward Models
Paper • 2609.32720 • Published • 25
Rubric Rewards from Item Response Theory
Paper • 2609.35646 • Published • 19
Sharpening Tax in Post-Training
Paper • 2610.01509 • Published • 96
On-Policy or Off-Policy Learning? A Systematic Study of Distillation Dynamics
Paper • 2609.35259 • Published • 188
Better Supervision Is Nearby: Neighborhood On-Policy Self-Distillation
Paper • 2609.39687 • Published • 18
Decoding Looped Transformers Better for (Almost) Free
Paper • 2610.02185 • Published • 43
upvoted 6
papers 5 days ago
Label-free steering: Compressing test-time reinforcement learning into bias-only subspaces
Paper • 2609.18587 • Published • 4
Synthetic Pre-pretraining Survives Scale, but Not as a Grammatical Prior
Paper • 2609.39827 • Published • 12
Scaling Laws for Looped Mixture of Experts
Paper • 2609.40316 • Published • 22
Overcoming Scaling Limits in On-Policy Self-Distillation for LLM Reasoning
Paper • 2609.37915 • Published • 8
Not Every Token Is Worth Distilling: Selective Supervision for Direct-OPD
Paper • 2609.29142 • Published • 5
The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation
Paper • 2609.36484 • Published • 578
upvoted 2
papers 8 days ago
upvoted a paper 10 days ago
upvoted 3
papers 11 days ago
Six Layers Less: Encoder Pruning for Whisper with Label-Free Recovery
Paper • 2609.27980 • Published • 6
Capable yet Parsimonious: Extracting and Characterizing Hidden Chain-of-Thought in Frontier Models
Paper • 2609.26637 • Published • 26
Your Transformer Can Hold Two Thoughts at Once: Evidence of Linear Superposition in LLMs
Paper • 2609.29845 • Published • 104