4 citations · 7 across the 12 of their papers we have counts for
12 papers
Efficient Mixture-of-Experts with Speculative Decoding via Expert Coactivation
Kumari Nishu, Han-Byul Kim, Santosh Chilkunda +13
Mixture-of-Experts (MoE) models are increasingly deployed alongside Speculative Decoding (SD) to accelerate inference, but combining the two is challenging. SD improves the inferen…
DACA-GRPO: Denoising-Aware Credit Assignment for Reinforcement Learning in Diffusion Language Models
Amin Karimi Monsefi, Dominic Culver, Nikhil Bhendawade +4
Diffusion large language models are a compelling alternative to autoregressive models, yet existing RL methods for diffusion treat all denoising steps as equally important and rely…
Trajectory as the Teacher: Few-Step Discrete Flow Matching via Energy-Navigated Distillation
Amin Karimi Monsefi, Dominic Culver, Nikhil Bhendawade +3
Discrete flow matching generates text by iteratively transforming noise tokens into coherent language, but may require hundreds of forward passes. Distillation uses the multi-step…
The Design Space of Tri-Modal Masked Diffusion Models
Louis Bethune, Victor Turrisi, Bruno Kacper Mlodozeniec +21
Discrete diffusion models have emerged as strong alternatives to autoregressive language models, with recent work initializing and fine-tuning a base unimodal model for bimodal gen…
Mirror Speculative Decoding: Breaking the Serial Barrier in LLM Inference
Nikhil Bhendawade, Kumari Nishu, Arnav Kundu +3
Speculative decoding accelerates LLM inference by using a draft model to look ahead, but gains are capped by the cost of autoregressive draft generation: increasing draft size elev…
FS-DFM: Fast and Accurate Long Text Generation with Few-Step Diffusion Language Models
Amin Karimi Monsefi, Nikhil Bhendawade, Manuel Rafael Ciosici +3
Autoregressive language models (ARMs) deliver strong likelihoods, but are inherently serial: they generate one token per forward pass, which limits throughput and inflates latency…
Apple Intelligence Foundation Language Models: Tech Report 2025
Ethan Li, Anders Boesen Lindbo Larsen, Chen Zhang +395
We introduce two multilingual, multimodal foundation language models that power Apple Intelligence features across Apple devices and services: i a 3B-parameter on-device model opti…
M2R2: Mixture of Multi-Rate Residuals for Efficient Transformer Inference
Nikhil Bhendawade, Mahyar Najibi, Devang Naik +1
Residual transformations enhance the representational depth and expressive power of large language models (LLMs). However, applying static residual transformations across all token…
Apple Intelligence Foundation Language Models
Tom Gunter, Zirui Wang, Chong Wang +152
We present foundation language models developed to power Apple Intelligence features, including a ~3 billion parameter model designed to run efficiently on devices and a large serv…
Speculative Streaming: Fast LLM Inference without Auxiliary Models
Nikhil Bhendawade, Irina Belousova, Qichen Fu +3
Speculative decoding is a prominent technique to speed up the inference of a large target language model based on predictions of an auxiliary draft model. While effective, in appli…
FastSeq: Make Sequence Generation Faster
Yu Yan, Fei Hu, Jiusheng Chen +7
Transformer-based models have made tremendous impacts in natural language generation. However the inference speed is a bottleneck due to large model size and intensive computing in…
EL-Attention: Memory Efficient Lossless Attention for Generation
Yu Yan, Jiusheng Chen, Weizhen Qi +4
Transformer model with multi-head attention requires caching intermediate results for efficient inference in generation tasks. However, cache brings new memory-related costs and pr…