arXiv:2606. 09514v1 Announce Type: new Abstract: Large language models (LLMs) incur high inference cost due to their depth and parameter scale.
arXiv:2606. 06467v1 Announce Type: cross Abstract: Long-context inference in modern LLMs is increasingly constrained by decoding efficiency, especially in reasoning-heavy settings where models generate long intermediate chains of thought.
MoRE: Mixture of Reused Experts is a hybrid architecture that combines Mixture-of-Experts (MoE) with weight‑sharing techniques. It shares expert pools across adjacent layers while each layer keeps its own router, and introduces lightweight depth embeddings to help shared experts differentiate layer contexts. Experiments on models ranging from 114 M to 1.15 B parameters show MoRE achieves lower perplexity and better downstream performance than standard MoEs and other weight‑sharing models, with only minimal changes to existing MoE implementations.
arXiv:2608. 05872v1 Announce Type: cross Abstract: Standard Large Language Models (LLMs) execute layers sequentially.
arXiv:2607. 06523v1 Announce Type: new Abstract: Long-context language model inference is increasingly limited by the memory bandwidth and capacity required to store key-value caches, yet existing compression methods often apply uniform budgets across layers or tokens and degrade retrieval when lexical cues and semantic states require different preservation.
arXiv:2606. 27743v1 Announce Type: cross Abstract: Large Language Models (LLMs) inference is typically deployed under a static resource assumption, where models execute a fixed computational graph regardless of the runtime environment.
T-LoopFormer introduces token-level elastic-depth looped transformers that allow each token to decide its own number of loop iterations based on its hidden state, improving token generation accuracy. It also adds a recursion-wise key‑value cache so tokens at different depths only attend to their corresponding cached states, speeding up autoregressive decoding. Experiments demonstrate strong performance on language modeling and zero‑shot reasoning, achieving the lowest decoding latency among comparable models.
The paper introduces Ban&Pick, a post‑training, plug‑and‑play routing strategy for Sparse Mixture‑of‑Experts large language models. It identifies and reinforces a small group of highly influential experts while dynamically pruning redundant ones, leading to accuracy gains across math, code, and reasoning benchmarks. Experiments on DeepSeek and Qwen3 show notable performance improvements and a 1.25× inference speedup without retraining or architectural changes.
arXiv:2609.08189v1 Announce Type: new Abstract: Dynamic layer routing reduces the inference cost of Large Language Models (LLMs) by learning to skip layers for individual tokens. Existing methods, ho...
arXiv:2510. 19366v2 Announce Type: replace-cross Abstract: Mixture-of-Experts (MoE) scales model capacity through sparse activation, and is becoming an important architecture for large language models (LLMs).
arXiv:2606. 09659v1 Announce Type: cross Abstract: Long-context language model inference is bottlenecked by memory, as the KV cache grows with context length.
arXiv:2608.28726v1 Announce Type: new Abstract: The remarkable performance of multimodal large language models (MLLMs) comes at the cost of substantial computational overhead, posing significant chal...
GroupKV is a lightweight hierarchical KV cache management system designed for long‑context diffusion large language model (dLLM) inference. It partitions the context into contiguous groups and uses coarse‑to‑fine sparse selection, cross‑layer consistency for predictive prefetching, and a staleness correction mechanism to keep the cache coherent amid dynamic KV updates. The approach also incorporates streaming prefill to lower peak memory usage, achieving up to 48× longer serviceable context, 3.73× faster inference in offload‑based settings, and competitive task accuracy.