Faster and Better LLMs via Latency-Aware Test-Time Scaling
Paper
• 2505.19634 • Published
Zili Wang
MarkWang
AI & ML interests
Multi-modality learning and inference acceleration
Organizations
authored a paper 9 months ago
authored 5 papers over 1 year ago
Continuous Speculative Decoding for Autoregressive Image Generation
Paper
• 2411.11925 • Published
• 16
Draw an Audio: Leveraging Multi-Instruction for Video-to-Audio Synthesis
Paper
• 2409.06135 • Published
• 16
Layerwise Recurrent Router for Mixture-of-Experts
Paper
• 2408.06793 • Published
• 32
AVESFormer: Efficient Transformer Design for Real-Time Audio-Visual Segmentation
Paper
• 2408.01708 • Published
• 4
A Closer Look into Mixture-of-Experts in Large Language Models
Paper
• 2406.18219 • Published
• 17