Pith. sign in

REVIEW 12 cited by

Higher Layers Need More LoRA Experts

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 2402.08562 v1 pith:W32PNKH6 submitted 2024-02-13 cs.CL cs.AI

Higher Layers Need More LoRA Experts

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

Parameter-efficient tuning (PEFT) techniques like low-rank adaptation (LoRA) offer training efficiency on Large Language Models, but their impact on model performance remains limited. Recent efforts integrate LoRA and Mixture-of-Experts (MoE) to improve the performance of PEFT methods. Despite promising results, research on improving the efficiency of LoRA with MoE is still in its early stages. Recent studies have shown that experts in the MoE architecture have different strengths and also exhibit some redundancy. Does this statement also apply to parameter-efficient MoE? In this paper, we introduce a novel parameter-efficient MoE method, \textit{\textbf{M}oE-L\textbf{o}RA with \textbf{L}ayer-wise Expert \textbf{A}llocation (MoLA)} for Transformer-based models, where each model layer has the flexibility to employ a varying number of LoRA experts. We investigate several architectures with varying layer-wise expert configurations. Experiments on six well-known NLP and commonsense QA benchmarks demonstrate that MoLA achieves equal or superior performance compared to all baselines. We find that allocating more LoRA experts to higher layers further enhances the effectiveness of models with a certain number of experts in total. With much fewer parameters, this allocation strategy outperforms the setting with the same number of experts in every layer. This work can be widely used as a plug-and-play parameter-efficient tuning approach for various applications. The code is available at https://github.com/GCYZSL/MoLA.

discussion (0)

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

Forward citations

Cited by 12 Pith papers

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

  1. A Data-Efficient Path to Multilingual LLMs: Language Expansion via Post-training PARAM$\Delta$ Integration into Upcycled MoE

    cs.CL 2026-05 unverdicted novelty 7.0

    PARAMΔ upcycles dense models to MoE for per-language experts and grafts post-training deltas to enable data-efficient language expansion while preserving original capabilities.

  2. 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.

  3. Behavioral and Representational Evidence of Binomial Ordering Preferences in Large Language Models

    cs.CL 2026-06 unverdicted novelty 6.0

    LLMs recover dominant binomial orders from corpora but align less closely with exact preference distributions, with preference strength partially encoded in middle-to-late layers and manipulable via steering.

  4. A$_3$B$_2$: Adaptive Asymmetric Adapter for Alleviating Branch Bias in Vision-Language Image Classification with Few-Shot Learning

    cs.CV 2026-05 unverdicted novelty 6.0

    A3B2 adds uncertainty-aware dampening and asymmetric MoE-style adapters to balance image and text branches, outperforming 11 baselines on 11 few-shot datasets.

  5. A$_3$B$_2$: Adaptive Asymmetric Adapter for Alleviating Branch Bias in Vision-Language Image Classification with Few-Shot Learning

    cs.CV 2026-05 unverdicted novelty 6.0

    A3B2 introduces an adaptive asymmetric adapter with uncertainty-aware dampening to reduce branch bias in few-shot vision-language image classification and outperforms standard adapter and prompt methods.

  6. TAS-LoRA: Transformer Architecture Search with Mixture-of-LoRA Experts

    cs.CV 2026-05 unverdicted novelty 6.0

    TAS-LoRA attaches a mixture of LoRA experts to a supernet and uses a dynamic router plus group-wise initialization to let different architecture subnets learn distinct features, yielding higher accuracy than prior TAS...

  7. CoMoL: Efficient Mixture of LoRA Experts via Dynamic Core Space Merging

    cs.CL 2026-02 conditional novelty 6.0

    CoMoL represents every LoRA expert as a shared-basis core matrix and merges token-selected experts in that core space, reaching standard LoRA parameter counts while outperforming MoE-LoRA baselines on math and code.

  8. FedTreeLoRA: Reconciling Statistical and Functional Heterogeneity in Federated LoRA Fine-Tuning

    cs.LG 2026-02 conditional novelty 6.0

    A federated LoRA fine-tuning method that builds a client-similarity tree and adapts aggregation depth layer-by-layer outperforms flat or global aggregation baselines on NLU and NLG tasks.

  9. LoRA-Mixer: Coordinate Modular LoRA Experts Through Serial Attention Routing

    cs.LG 2025-06 unverdicted novelty 6.0

    LoRA-Mixer routes modular LoRA experts into attention projection matrices with an adaptive Routing Specialization Loss to improve multi-task performance while using fewer trainable parameters than prior LoRA-MoE methods.

  10. Curvature-Weighted Capacity Allocation: A Minimum Description Length Framework for Layer-Adaptive Large Language Model Optimization

    cs.IT 2026-03 conditional novelty 4.0

    A convex-programming framework with closed-form water-filling solutions converts layer curvature scores into optimal expert allocation and pruning ratios for LLMs, with a quadratic transfer-regret bound.

  11. Curvature-Weighted Capacity Allocation: A Minimum Description Length Framework for Layer-Adaptive Large Language Model Optimization

    cs.IT 2026-03 reject novelty 3.0

    A curvature-based layer gain is converted into optimal allocation and pruning via water-filling and convex penalties, but the experiments substitute the baseline's influence scores and show only modest, mixed gains.

  12. DP-FedLoRA: Privacy-Enhanced Federated Fine-Tuning for On-Device Large Language Models

    cs.CR 2025-09 reject novelty 3.0

    DP-FedLoRA clips and adds Gaussian noise to per-client LoRA matrices in federated LLM fine-tuning, claiming unbiased updates and bounded variance, but the privacy calibration and experiments have significant gaps.