Pith. sign in

REVIEW 10 cited by

LoRA-XS: Low-Rank Adaptation with Extremely Small Number of Parameters

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 2405.17604 v3 pith:S7TZDDAR submitted 2024-05-27 cs.LG cs.AIcs.CL

LoRA-XS: Low-Rank Adaptation with Extremely Small Number of Parameters

classification cs.LG cs.AIcs.CL
keywords lora-xslargeloramodelsparametersstoragetrainablecomputational
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

The growth of large language models underscores the need for parameter-efficient fine-tuning. Despite its popularity, LoRA encounters storage and computational challenges when deploying multiple task- or user-specific modules. To address this, we introduce LoRA-XS, a novel fine-tuning method backed by a theoretical derivation. LoRA-XS drastically reduces trainable parameters by incorporating a small, trainable weight matrix between frozen low-rank matrices derived from the Singular Value Decomposition of pre-trained weights. This design enables LoRA-XS to reduce storage requirements by over 100x in 7B models compared to LoRA. Additionally, unlike other methods, LoRA-XS imposes no lower bound on trainable parameters - it can scale from a single parameter per module to arbitrarily large values, adapting to any storage or computational constraint. Evaluations on GLUE, GSM8K, MATH, and commonsense reasoning benchmarks across different model scales reveal that LoRA-XS consistently outperforms or matches LoRA and VeRA in accuracy, offering unmatched parameter efficiency. Our ablation studies highlight the significance of singular vectors in transformer weights, establishing LoRA-XS as a powerful, storage-efficient solution for scaling and personalizing large language models.

discussion (0)

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

Forward citations

Cited by 10 Pith papers

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

  1. LOFT: Low-Rank Orthogonal Fine-Tuning via Task-Aware Support Selection

    cs.LG 2026-05 unverdicted novelty 7.0

    LOFT unifies orthogonal PEFT by treating adaptation as low-rank subspace rotation and adds task-aware support selection that improves efficiency under fixed budgets.

  2. BoostLoRA: Growing Effective Rank by Boosting Adapters

    cs.LG 2026-04 unverdicted novelty 7.0

    BoostLoRA grows effective adapter rank linearly via iterative boosting on hard examples with orthogonal low-rank updates, outperforming both single-shot ultra-low-rank adapters and full fine-tuning on math and code ta...

  3. Spectral Rewiring for Exploration, Purification, and Model Merging

    cs.LG 2026-07 conditional novelty 6.5

    Subspace-Aligned Rewiring projects RL weight updates onto the base model’s SVD basis, retaining a compact rewiring matrix that preserves reasoning and improves exploration and multi-domain merging.

  4. Non-vacuous Generalization Bounds for Reinforcement Learning with Verifiable Rewards

    cs.LG 2026-07 reject novelty 6.0

    Non-vacuous PAC-Bayes generalization bounds for billion-parameter RLVR models, obtained by a Gumbel-max reparameterization and aggressive TinyLoRA distillation/quantization, are claimed for four tasks.

  5. Bayesian Fine-tuning in Projected Subspaces

    cs.LG 2026-05 unverdicted novelty 6.0

    Bayesian fine-tuning of large models can be done efficiently by projecting uncertainties into low-dimensional subspaces, yielding improved calibration and generalization while keeping computational costs low.

  6. SOS-LoRA: Static Orthogonal-Subspace Low-Rank Adaptation with Fixed Multi-Scale Scaling

    cs.LG 2026-06 conditional novelty 5.0

    A LoRA update split into several fixed, differently-scaled low-rank experts with orthogonal input directions improves fine-tuning accuracy at the same parameter count.

  7. The Hidden Power of Scaling Factor in LoRA Optimization

    cs.AI 2026-06 unverdicted novelty 5.0

    Alpha in LoRA outperforms learning-rate scaling, follows a square-root law with rank, and enables a minimalist LoRA-alpha method that improves performance across tasks.

  8. LoCO: Low-rank Compositional Rotation Fine-tuning

    cs.LG 2026-05 unverdicted novelty 5.0

    LoCO is a PEFT technique that constructs orthogonal transformations via low-rank skew-symmetric matrices and compositional rotation chains with a parallelizable approximation, validated on transformer adaptations.

  9. CERSA: Cumulative Energy-Retaining Subspace Adaptation for Memory-Efficient Fine-Tuning

    cs.LG 2026-05 unverdicted novelty 5.0

    CERSA derives low-rank fine-tuning subspaces from SVD principal components that retain 90-95% spectral energy, delivering higher performance than LoRA and other PEFT baselines at substantially lower memory cost across...

  10. GiVA: Gradient-Informed Bases for Vector-Based Adaptation

    cs.CL 2026-04 unverdicted novelty 5.0

    GiVA uses gradients to initialize vector adapters so they match LoRA performance at eight times lower rank while keeping extreme parameter efficiency.