Efficient Pre-Training with Token Superposition
Paper • 2605.06546 • Published • 44
Stlanik
Stlanik
·
AI & ML interests
None yet
Recent Activity
liked a model 2 days ago
PsycheFoundation/dm-fwedu-dclm liked a model 2 days ago
PsycheFoundation/moe-10b-a1b-init upvoted a paper 2 days ago
Efficient Pre-Training with Token SuperpositionOrganizations
None yet
upvoted 4
papers 2 days ago
Long Context Pre-Training with Lighthouse Attention
Paper • 2605.06554 • Published • 29
Targeted Neuron Modulation via Contrastive Pair Search
Paper • 2605.12290 • Published • 14
Decoupling the Benefits of Subword Tokenization for Language Model Training via Byte-level Simulation
Paper • 2604.27263 • Published • 9