# Enhancing Multimodal Continual Instruction Tuning with BranchLoRA

arXiv: https://arxiv.org/abs/2506.02041 · HTML: https://mlsanity.com/paper/2506.02041/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- Placed in Prompt learning and adaptation of CLIP, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 24 of its 49 references that are in the map.
- First posted on arXiv 31 May 2025.

## Placed in

- Multimodal LLMs / vision-language models / [Prompt learning and adaptation of CLIP](https://mlsanity.com/mllm/prompt-learning-and-adaptation-of-clip/)
- Mixture-of-experts / [Mixture of LoRA experts](https://mlsanity.com/moe/mixture-of-lora-experts/)

## Builds on (most cited first, 24 of 24)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://mlsanity.com/paper/2101.03961/) (2021)
- [InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning](https://mlsanity.com/paper/2305.06500/) (2023)
- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [Learning without Forgetting](https://mlsanity.com/paper/1606.09282/) (2016)
- [Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering](https://mlsanity.com/paper/2209.09513/) (2022)
- [Towards VQA Models That Can Read](https://mlsanity.com/paper/1904.08920/) (2019)
- [ShareGPT4V: Improving Large Multi-Modal Models with Better Captions](https://mlsanity.com/paper/2311.12793/) (2023)
- [Learning to Prompt for Continual Learning](https://mlsanity.com/paper/2112.08654/) (2021)
- [Generation and Comprehension of Unambiguous Object Descriptions](https://mlsanity.com/paper/1511.02283/) (2015)
- [VizWiz Grand Challenge: Answering Visual Questions from Blind People](https://mlsanity.com/paper/1802.08218/) (2018)
- [An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networks](https://mlsanity.com/paper/1312.6211/) (2013)
- [From System 1 to System 2: A Survey of Reasoning Large Language Models](https://mlsanity.com/paper/2502.17419/) (2025)
- [Patching open-vocabulary models by interpolating weights](https://mlsanity.com/paper/2208.05592/) (2022)
- [VLP: A Survey on Vision-Language Pre-training](https://mlsanity.com/paper/2202.09061/) (2022)
- [ChartCoder: Advancing Multimodal Large Language Model for Chart-to-Code Generation](https://mlsanity.com/paper/2501.06598/) (2025)
- [Lifelong Learning of Large Language Model based Agents: A Roadmap](https://mlsanity.com/paper/2501.07278/) (2025)
- [Model Merging by Uncertainty-Based Gradient Matching](https://mlsanity.com/paper/2310.12808/) (2023)
- [Continual Instruction Tuning for Large Multimodal Models](https://mlsanity.com/paper/2311.16206/) (2023)
- [Mitigating Catastrophic Forgetting in Language Transfer via Model Merging](https://mlsanity.com/paper/2407.08699/) (2024)
- [ChartEdit: How Far Are MLLMs From Automating Chart Analysis? Evaluating MLLMs' Capability via Chart Editing](https://mlsanity.com/paper/2505.11935/) (2025)
- [Beyond Anti-Forgetting: Multimodal Continual Instruction Tuning with Positive Forward Transfer](https://mlsanity.com/paper/2401.09181/) (2024)
- [DEM: Distribution Edited Model for Training with Mixed Data Distributions](https://mlsanity.com/paper/2406.15570/) (2024)

## Built on by (most cited first, 14 of 14)

- [Continual Learning for VLMs: A Survey and Taxonomy Beyond Forgetting](https://mlsanity.com/paper/2508.04227/) (2025)
- [Continual Learning for Generative AI: From LLMs to MLLMs and Beyond](https://mlsanity.com/paper/2506.13045/) (2025)
- [PASs-MoE: Mitigating Misaligned Co-drift among Router and Experts via Pathway Activation Subspaces for Continual Learning](https://mlsanity.com/paper/2601.13020/) (2026)
- [Harmonious Parameter Adaptation in Continual Visual Instruction Tuning for Safety-Aligned MLLMs](https://mlsanity.com/paper/2511.20158/) (2025)
- [Lifelong Embodied Navigation Learning](https://mlsanity.com/paper/2603.06073/) (2026)
- [All-day Multi-scenes Lifelong Vision-and-Language Navigation with Tucker Adaptation](https://mlsanity.com/paper/2603.14276/) (2026)
- [On Token's Dilemma: Dynamic MoE with Drift-Aware Token Assignment for Continual Learning of Large Vision Language Models](https://mlsanity.com/paper/2603.27481/) (2026)
- [FTibSuite: A Comprehensive Resource Suite for Tibetan Vision-Language Modeling](https://mlsanity.com/paper/2605.26601/) (2026)
- [Attention-Spectrum Regularization for Replay-Free Continual Multimodal LLMs](https://mlsanity.com/paper/2606.23063/) (2026)
- [Hidden Forgetting in Continual Multimodal Learning: When Accuracy Survives but Grounding Fails](https://mlsanity.com/paper/2607.02020/) (2026)
- [StrLoRA: Towards Streaming Continual Visual Instruction Tuning for MLLMs](https://mlsanity.com/paper/2605.16353/) (2026)
- [InduceKV: Fixed-Footprint Continual Adaptation of Multimodal LLMs via Inducing KV Memories](https://mlsanity.com/paper/2607.02010/) (2026)
- [Progressive Multimodal Alignment for Continual Instruction Tuning](https://mlsanity.com/paper/2607.26947/) (2026)
- [Hyper-LLaVA: Hyperbolic Uncertainty-aware Modality-Balanced Routing for Multimodal Continual Instruction Tuning](https://mlsanity.com/paper/2609.13742/) (2026)

## Abstract (arXiv)

Multimodal Continual Instruction Tuning (MCIT) aims to finetune Multimodal Large Language Models (MLLMs) to continually align with human intent across sequential tasks. Existing approaches often rely on the Mixture-of-Experts (MoE) LoRA framework to preserve previous instruction alignments. However, these methods are prone to Catastrophic Forgetting (CF), as they aggregate all LoRA blocks via simple summation, which compromises performance over time. In this paper, we identify a critical parameter inefficiency in the MoELoRA framework within the MCIT context. Based on this insight, we propose BranchLoRA, an asymmetric framework to enhance both efficiency and performance. To mitigate CF, we introduce a flexible tuning-freezing mechanism within BranchLoRA, enabling branches to specialize in intra-task knowledge while fostering inter-task collaboration. Moreover, we incrementally incorporate task-specific routers to ensure an optimal branch distribution over time, rather than favoring the most recent task. To streamline inference, we introduce a task selector that automatically routes test inputs to the appropriate router without requiring task identity. Extensive experiments on the latest MCIT benchmark demonstrate that BranchLoRA significantly outperforms MoELoRA and maintains its superiority across various MLLM sizes.