arXiv:2608. 12385v1 Announce Type: new Abstract: As large language models serve more requests, cumulative inference cost is becoming increasingly important relative to one-time training cost.
arXiv:2608. 12385v2 Announce Type: replace Abstract: As large language models serve ever more requests, cumulative inference cost is growing relative to the one-time cost of training.
arXiv:2608. 08888v1 Announce Type: new Abstract: Autoregressive transformers compute along two axes: horizontally across generated tokens, and vertically through model depth.
arXiv:2606. 15453v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) based large language models (LLMs), such as Qwen and DeepSeek, have recently emerged as an effective approach to improving model capacity without proportionally increasing computational cost.
arXiv:2607. 06601v1 Announce Type: cross Abstract: Conditional computation can decouple language model quality from per-token inference cost, yet leading techniques act on a single axis in isolation: Mixture-of-Experts (MoE) sparsifies the FFN, Mixture-of-Depths (MoD) skips whole transformer blocks, and KV-cache quantization compresses attention memory.
arXiv:2607. 24434v1 Announce Type: cross Abstract: Large Mixture-of-Experts (MoE) language models are attractive for end-device deployment because only a small subset of experts is active per token, but their routed expert weights often exceed accelerator memory.
arXiv:2607. 12696v1 Announce Type: cross Abstract: Sparse Mixture-of-Experts (MoE) models have become an important approach for scaling Large Language Models (LLMs), but their inference efficiency depends strongly on expert activation patterns.
Large Mixture-of-Experts (MoE) language models are attractive for end-device deployment because only a small subset of experts is active per token, but their routed expert weights often exceed accelerator memory. We target latency-critical single-user settings where routed experts are staged on demand from CPU memory to a GPU or from Flash to a mobile NPU.
arXiv:2607. 27600v1 Announce Type: new Abstract: Key-value (KV) cache management through compression and eviction strategies has emerged as an important research direction in recent years.
arXiv:2511. 21759v2 Announce Type: replace-cross Abstract: Diffusion-based large language models (dLLMs) have recently gained significant attention for their exceptional performance and inherent potential for parallel decoding.
arXiv:2607. 01831v1 Announce Type: cross Abstract: Long-context inference is increasingly common in large language model (LLM) serving, driven by retrieval-augmented generation and agentic systems.
arXiv:2607. 24260v1 Announce Type: new Abstract: Modern LLM systems increasingly rely on knowledge-selection processes that produce high-value structured priors, such as ranked evidence, graph topology, multimodal alignment, and confidence signals.
Long chain-of-thought (CoT) trajectories in large language model (LLM) reasoning cause severe inference bottlenecks due to rapid key-value (KV) cache growth. Current decoding-time compression methods mitigate this issue via token eviction, but typically assume a uniform budget distribution across all layers and heads.