LayerRoute is a parameter‑efficient technique that enables adaptive skipping of transformer layers in large language models. It adds a lightweight per‑layer router (~21.5K parameters) and LoRA adapters (rank 8, ~1.08M parameters) to each of the 24 blocks in Qwen2.5‑0.5B‑Instruct, training them jointly with a gate‑regularized language‑modeling objective. Across ten independent runs, the method consistently identifies nine middle layers as skip‑eligible, achieves a verified wall‑clock speedup of 1.02x–1.06x, and improves perplexity by an average of 1.16 points, while the router’s decisions vary per input, confirming genuine adaptive behavior.
By Prateek Kumar Sikdar
arXiv:2606. 01838v1 Announce Type: cross Abstract: Agentic language model systems alternate between two structurally distinct step types: structured tool calls (short, deterministic, low perplexity) and open-ended planning/reasoning steps (long, complex, high perplexity).
By Prateek Kumar Sikdar
arXiv:2608. 16391v1 Announce Type: cross Abstract: As large language models become increasingly widespread, third-party providers that deploy open-weight models have become an important part of the ecosystem.
By Xiangfan Wu, Zonghao Ying, Huiyu Wu, Xing Zheng, Huangsheng Cheng, Xiaorong Shi, Jing Guo
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.
By Dengke Han
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:2608. 03839v1 Announce Type: new Abstract: Training-free speculative decoding drafts by matching an exact suffix of the context against a pool of earlier context.
By Tao Jin, Phuong Minh Nguyen, Zhenzhu Yan, Teeradaj Racharak, Naoya Inoue
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.
By Yuanteng Chen, Peisong Wang, Yuantian Shao, Nanxin Zeng, Chang Xu, Jian Cheng
arXiv:2606. 12243v1 Announce Type: cross Abstract: Speculative decoding (SD) addresses the high inference costs of LLMs by having lightweight drafters generate candidates for large verifiers to validate in parallel.
By Yuchen Xian, Yang He, Yunqiu Xu, Yi Yang
arXiv:2608. 08853v1 Announce Type: new Abstract: Sparse Mixture-of-Experts (MoE) routers commonly use the same scores both to select experts and to weight their already-computed outputs.
By Zongfei Li
arXiv:2605. 17106v2 Announce Type: replace-cross Abstract: Production LLM deployments increasingly maintain heterogeneous model pools spanning order-of-magnitude cost differences.
By Aashna Garg, Siddharth Singha Roy, Jinu Jang, Federico Brancasi, Shengyu Fu
The paper introduces Untied Self-Conditioning, a sampler that corrects a train–inference mismatch in flow‑matching language models. By dampening redundant directions in the self‑conditioning input and approximating a step‑average prediction from history, the method improves generation quality without retraining. On LangFlow and ELF‑B datasets, it dramatically lowers perplexity and is preferred in the majority of pairwise comparisons.
By Bocheng Li, Linli Xu
Every memory-based knowledge editor in the SERAC lineage depends on a scope decision: given a query, does a stored edit apply? We report that current knowledge-editing benchmarks cannot measure this d...