StageML, a partial evaluation (PE) system that exploits the temporal boundaries between deployment, adapter loading, tenant request admission, expert routing, and token generation using a binding-time lattice, and the results show that PE complements low-level kernel optimization.
This work introduces RouteFM, which learns to characterize anonymous candidate models from behavioral context and infer their target-specific capabilities, rather than binding routing decisions to fixed model identities or a single environment, and shows transfer across changes in domains, modalities, candidate pools,...
DynaNDE is presented, a dynamic near-data expert scheduling framework that exploits NPU-NDP collaboration to accelerate batched MoE inference and introduces an analytical performance model that captures hardware heterogeneity, data-movement costs, and communication-computation overlap in cooperative NPU-NDP execution.
Results characterize when runtime-controlled composition of coding-agent products improves execution accuracy, as well as when runtime-controlled composition of coding-agent products improves execution accuracy.
Zheng-Yu Chen, Lin-Feng Liu, Hong Li et al.· 0 citations
OptiCom is proposed, a unified framework that represents LLM-driven optimizers within a shared configuration space: C=(A,Q,O,E,M,S), corresponding to artifact, query, operator, evaluation, memory, and strategy, corresponding to artifact, query, operator, evaluation, memory, and strategy.
Chen-Xing Wei, Si-Chen Liu, Lizzie Liu et al.· 0 citations
Across diverse tasks and roles, model scales, and MoE architectures, DEP achieves better overall accuracy than static pruning and merging baselines, and generalizes to workflows unseen in training without retraining.
Jabin Koo, Soheil Abbasloo, Sungjae Lee et al.· 0 citations
Nereus is a cost-aware runtime that adapts RL post-training jobs into efficient execution plans and executes a transition using a memory-feasible global plan and admits the transition using a cost model calibrated against the running job.
Songlin Jiang, Tuo Shi, Si-Tong Zhang et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
