# Self-Routed Tensor Adapters for Parameter-Efficient Universal Visual Adaptation

arXiv: https://arxiv.org/abs/2608.16384 · HTML: https://mlsanity.com/paper/2608.16384/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Vision mixture-of-experts, one of 12 directions in Mixture-of-experts.
- Builds on 20 of its 24 references that are in the map.
- First posted on arXiv 17 Aug 2026.

## Placed in

- Mixture-of-experts / [Vision mixture-of-experts](https://mlsanity.com/moe/vision-mixture-of-experts/)

## Builds on (most cited first, 20 of 20)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Prefix-Tuning: Optimizing Continuous Prompts for Generation](https://mlsanity.com/paper/2101.00190/) (2021)
- [The Power of Scale for Parameter-Efficient Prompt Tuning](https://mlsanity.com/paper/2104.08691/) (2021)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://mlsanity.com/paper/2101.03961/) (2021)
- [Visual Prompt Tuning](https://mlsanity.com/paper/2203.12119/) (2022)
- [Deep Hashing Network for Unsupervised Domain Adaptation](https://mlsanity.com/paper/1706.07522/) (2017)
- [Deeper, Broader and Artier Domain Generalization](https://mlsanity.com/paper/1710.03077/) (2017)
- [DoRA: Weight-Decomposed Low-Rank Adaptation](https://mlsanity.com/paper/2402.09353/) (2024)
- [Deeply-Supervised Nets](https://mlsanity.com/paper/1409.5185/) (2014)
- [VeRA: Vector-based Random Matrix Adaptation](https://mlsanity.com/paper/2310.11454/) (2023)
- [Deep Domain-Adversarial Image Generation for Domain Generalisation](https://mlsanity.com/paper/2003.06054/) (2020)
- [Pushing Mixture of Experts to the Limit: Extremely Parameter Efficient MoE for Instruction Tuning](https://mlsanity.com/paper/2309.05444/) (2023)
- [FacT: Factor-Tuning for Lightweight Adaptation on Vision Transformer](https://mlsanity.com/paper/2212.03145/) (2022)
- [When MOE Meets LLMs: Parameter Efficient Fine-tuning for Multi-task Medical Applications](https://mlsanity.com/paper/2310.18339/) (2023)
- [NICO++: Towards Better Benchmarking for Domain Generalization](https://mlsanity.com/paper/2204.08040/) (2022)
- [VB-LoRA: Extreme Parameter Efficient Fine-Tuning with Vector Banks](https://mlsanity.com/paper/2405.15179/) (2024)
- [MALoRA: Mixture of Asymmetric Low-Rank Adaptation for Enhanced Multi-Task Learning](https://mlsanity.com/paper/2410.22782/) (2024)

## Built on by (most cited first, 1 of 1)

- [Score the Update, Not the Token: Descent-Aligned Routing for Combinatorial LoRA Experts](https://mlsanity.com/paper/2610.00493/) (2026)

## Abstract (arXiv)

Universal visual representations require adaptation mechanisms that adapt across heterogeneous domains without fragmenting knowledge into domain-specific modules. Parameter-efficient fine-tuning adapts frozen visual foundation models efficiently, but standard low-rank adapters use a fixed subspace for all inputs, which can be restrictive when domains differ in style, background, and semantic context. MoE-based adapters improve specialization through multiple expert pathways, but often rely on external routers and large expert banks, adding parameters and separating routing from adaptation. We propose Self-Routed Tensor Adapters, a compact framework for multi-domain visual adaptation. SRTA projects each input into a low-rank space, computes routing weights from this representation using a learnable domain matrix, and uses these weights to blend slices of a shared Tucker core. This produces a sample-specific adaptation matrix without an external gating network, allowing shared visual factors to be reused while supporting domain-aware specialization. To strengthen pathway learning, we introduce a progressive depth-weighted routing objective that supervises routing decisions across adapter layers. Across five heterogeneous multi-domain visual classification benchmarks, SRTA achieves competitive or slightly stronger average accuracy than MoE-style PEFT baselines while using substantially fewer trainable parameters. At rank 64, SRTA uses 2.77M parameters in the 4-domain setting compared with 9.52M for MoLoRA, and 3.00M in the 6-domain setting compared with 14.31M. Overall, SRTA offers an effective accuracy-parameter trade-off for adapting visual foundation models toward universal multi-domain representations. hrefhttps://github.com/surajyadav-research/SRTAGitHub