Pith. sign in

REVIEW 5 cited by

DA-MoE: Towards Dynamic Expert Allocation for Mixture-of-Experts Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2409.06669 v1 pith:ZERXMQ53 submitted 2024-09-10 cs.LG

DA-MoE: Towards Dynamic Expert Allocation for Mixture-of-Experts Models

classification cs.LG
keywords expertsmechanismmodelsroutertokenda-moedynamicimportance
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Transformer-based Mixture-of-Experts (MoE) models have been driving several recent technological advancements in Natural Language Processing (NLP). These MoE models adopt a router mechanism to determine which experts to activate for routing input tokens. However, existing router mechanisms allocate a fixed number of experts to each token, which neglects the varying importance of different input tokens. In this study, we propose a novel dynamic router mechanism that Dynamically Allocates a variable number of experts for Mixture-of-Experts (DA-MoE) models based on an effective token importance measure. First, we show that the Transformer attention mechanism provides a natural and effective way of calculating token importance. Second, we propose a dynamic router mechanism that effectively decides the optimal number of experts (K) and allocates the top-K experts for each input token. Third, comprehensive experiments on several benchmark datasets demonstrate that our DA-MoE approach consistently outperforms the state-of-the-art Transformer based MoE model on the popular GLUE benchmark.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Adaptive and Fine-grained Module-wise Expert Pruning for Efficient LoRA-MoE Fine-Tuning

    cs.LG 2026-04 unverdicted novelty 7.0

    DMEP prunes experts module-by-module in LoRA-MoE and removes load balancing after pruning, cutting trainable parameters 35-43% and raising throughput ~10% while matching or exceeding uniform baselines on reasoning tasks.

  2. ProbMoE: Differentiable Probabilistic Routing for Mixture-of-Experts

    cs.LG 2026-06 unverdicted novelty 6.0

    ProbMoE frames MoE routing as probabilistic inference over cardinality-constrained subsets, enabling Exact-k sampling with marginal-probability gradients and a dynamic-k variant that matches training and inference car...

  3. dMoE: dLLMs with Learnable Block Experts

    cs.CL 2026-05 unverdicted novelty 6.0

    dMoE aggregates token expert distributions to block level in dLLMs, cutting unique experts from 69.5 to 14.6, memory by 76-80%, and latency by 1.14-1.66x while retaining 99.11% performance.

  4. BEAM: Binary Expert Activation Masking for Dynamic Routing in MoE

    cs.AI 2026-05 conditional novelty 6.0

    BEAM uses binary expert activation masks trained end-to-end to achieve dynamic sparsity in MoE models, cutting FLOPs by 85% with over 98% performance retention.

  5. Beyond Uniform Experts: Cost-Aware Expert Execution for Efficient Multi-Device MoE Inference

    cs.DC 2026-06 unverdicted novelty 5.0

    CAEE reduces MoE inference latency 8-18% on 671B DeepSeek-R1 by cost-aware expert pruning and low-overhead compensation while keeping accuracy drop under 1%.