Less Is More: Reducing Token Counts Without Compromising Performance
arXiv:2506.15138v2 Announce Type: replace Abstract: Tokenization directly affects the inference efficiency of large language models, since fragmented tokenization increases sequence length and…
Read the full articleYou might also wanna read
LK Losses: A New Training Objective to Optimize Acceptance Rate in Speculative Decoding for LLMs
Speculative decoding accelerates autoregressive large language model (LLM) inference by using a lightweight draft model to propose candidate
arxiv.org·1mo ago
Fast-dLLM: Training-Free Acceleration Method for Diffusion Language Models Using KV Cache and Parallel Decoding
Diffusion-based large language models (Diffusion LLMs) have shown promise for non-autoregressive text generation with parallel decoding capa
arxiv.org·8mo ago
GenART: An adaptive genomic language model that dynamically discovers biologically meaningful DNA words
Chen et al. develop GenART, a genomic language model that dynamically segments raw DNA into biologically meaningful “words.” This adaptive a
Systematic Evaluation of Deep Learning Optimizers Reveals Limited Speedup Over AdamW in Language Model Pretraining
AdamW has long been the dominant optimizer in language model pretraining, despite numerous claims that alternative optimizers offer 1.4 to 2
arxiv.org·10mo ago
PHOTON: Hierarchical Autoregressive Model for Efficient Language Generation
Transformers operate as horizontal token-by-token scanners; at each generation step, the model attends to an ever-growing sequence of token-
arxiv.org·6mo ago
An Algorithm for Computing Optimal Tokenizers in Practice
In this post, I will present an algorithm that was able to compute an optimal tokenizer in some settings. This result is cool because optima
Comments
Sign in to join the conversation.
No comments yet. Be the first.