Pith. sign in

REVIEW 15 cited by

Mixture-of-Experts with Expert Choice Routing

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 2202.09368 v2 pith:THGSX5X4 submitted 2022-02-18 cs.LG cs.AI

classification cs.LGcs.AI
keywords expertexpertsmethodmixture-of-expertsnumbertokentokenstop-k
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Sparsely-activated Mixture-of-experts (MoE) models allow the number of parameters to greatly increase while keeping the amount of computation for a given token or a given sample unchanged. However, a poor expert routing strategy (e.g. one resulting in load imbalance) can cause certain experts to be under-trained, leading to an expert being under or over-specialized. Prior work allocates a fixed number of experts to each token using a top-k function regardless of the relative importance of different tokens. To address this, we propose a heterogeneous mixture-of-experts employing an expert choice method. Instead of letting tokens select the top-k experts, we have experts selecting the top-k tokens. As a result, each token can be routed to a variable number of experts and each expert can have a fixed bucket size. We systematically study pre-training speedups using the same computational resources of the Switch Transformer top-1 and GShard top-2 gating of prior work and find that our method improves training convergence time by more than 2x. For the same computational cost, our method demonstrates higher performance in fine-tuning 11 selected tasks in the GLUE and SuperGLUE benchmarks. For a smaller activation cost, our method outperforms the T5 dense model in 7 out of the 11 tasks.

Discussion (0). Sign in to comment.

Forward citations

Cited by 15 Pith papers

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

  1. Language-Assisted Super-Resolution from Real-World Low-Resolution Patches

    cs.CV 2026-06 unverdicted novelty 7.0 of 10

    LA-SR redefines unpaired super-resolution in language space by projecting images into a semantically rich representation and applying vision-language model guided losses to handle real-world degradations extracted fro...

  2. A Minimal Bifurcation Model of Load Imbalance in a Softmax Mixture-of-Experts Router

    math.DS 2026-05 unverdicted novelty 7.0 of 10

    A mean-field limit of a reinforcement-based softmax router for two experts shows a supercritical pitchfork bifurcation, with an external asymmetry unfolding it into a cusp of fold bifurcations.

  3. Affinity Is Not Enough: Recovering the Free Energy Principle in Mixture-of-Experts

    cs.LG 2026-05 conditional novelty 7.0 of 10

    Adding temporal memory via LIF, precision-weighted gating, and anticipatory prediction to MoE routers recovers effective expert selection at distribution transitions, with ablation confirming a super-additive beta-ant...

  4. Auxiliary-Loss-Free Load Balancing Strategy for Mixture-of-Experts

    cs.LG 2024-08 conditional novelty 7.0 of 10

    Loss-Free Balancing keeps expert loads balanced in MoE models by dynamically adjusting routing-score biases based on recent usage, avoiding auxiliary-loss interference and yielding better performance.

  5. Generic Expert Coverage for Pruning SparseMixture-of-Experts Language Models

    cs.AI 2026-07 unverdicted novelty 6.0 of 10

    Coverage-aware pruning using per-corpus utility profiles on WikiText2 and C4 improves zero-shot accuracy and reduces perplexity degradation in two MoE models at 25-75% retention compared to baselines, without downstream data.

  6. Schedule-Level Shared-Prefix Reuse for LLM RL Training

    cs.DC 2026-05 unverdicted novelty 6.0 of 10

    Schedule-level shared-prefix reuse decouples prefix and suffix passes in GRPO training to compute shared prefixes once, delivering up to 4.395x speedup and 59.1% HBM reduction while preserving numerical equivalence.

  7. Piper: Efficient Large-Scale MoE Training via Resource Modeling and Pipelined Hybrid Parallelism

    cs.DC 2026-05 unverdicted novelty 6.0 of 10

    Piper introduces resource modeling and pipelined hybrid parallelism for MoE training, delivering 2-3.5X higher MFU than prior frameworks and 1.2-9X better all-to-all bandwidth.

  8. AGoQ: Activation and Gradient Quantization for Memory-Efficient Distributed Training of LLMs

    cs.CL 2026-05 unverdicted novelty 6.0 of 10

    AGoQ delivers up to 52% lower memory use and 1.34x faster training for 8B-32B LLaMA models by using near-4-bit adaptive activations and 8-bit gradients while preserving pretraining convergence and downstream accuracy.

  9. HierMoE: Accelerating MoE Training with Hierarchical Token Deduplication and Expert Swap

    cs.DC 2025-08 conditional novelty 6.0 of 10

    HierMoE reduces MoE training time by removing duplicate token copies at each GPU-hierarchy level and swapping experts for load balance, measured at 1.18-1.27x end-to-end speedup on 32 GPUs.

  10. Mixture-of-Transformers: A Sparse and Scalable Architecture for Multi-Modal Foundation Models

    cs.CL 2024-11 conditional novelty 6.0 of 10

    MoT decouples non-embedding parameters by modality in transformers to match dense multi-modal performance with roughly one-third to one-half the FLOPs.

  11. Language-Assisted Super-Resolution from Real-World Low-Resolution Patches

    cs.CV 2026-06 unverdicted novelty 5.0 of 10

    LA-SR extracts real LR patches from depth-varying regions in single images and uses vision-language models with linguistic content and quality losses for unpaired super-resolution.

  12. AGoQ: Activation and Gradient Quantization for Memory-Efficient Distributed Training of LLMs

    cs.CL 2026-05 unverdicted novelty 5.0 of 10

    AGoQ cuts LLM training memory by up to 52% and speeds it up by 1.34x using tailored 4-bit activations and 8-bit gradients with special communication, matching baseline accuracy on LLaMA models.

  13. Maximum Score Routing For Mixture-of-Experts

    cs.LG 2025-08 unverdicted novelty 5.0 of 10

    MaxScore casts MoE routing as min-cost max-flow with SoftTopk and claims better loss and eval scores at equal FLOPs; unverified because the full text is unreadable.

  14. Neural Inhibition Improves Dynamic Routing and Mixture of Experts

    cs.LG 2025-07 conditional novelty 5.0 of 10

    Neural inhibition gating on MoE router inputs improves a synthetic digit/squares benchmark by about four points over plain MoE, but the language-model evidence is unreliable.

  15. Hecto: Modular Sparse Experts for Adaptive and Interpretable Reasoning

    cs.AI 2025-06 reject novelty 2.0 of 10

    A lightweight heterogeneous MoE with a GRU and an FFNN expert trails homogeneous baselines, and its claimed reasoning-type specialization is confounded by unequal expert inputs.

Pith tools