Challenge: Recent studies have explored Continual Instruction Tuning (CIT) in Multimodal Large Language Models (MLLMs), with a primary focus on Task-incremental CIT, where MLLM are required to continuously acquire new tasks.
Approach: They propose a Sparse Mixture of Expert (SMoE) based method for domain-incremental CIT in Multimodal Large Language Models (MLLMs) . they equip the SMoA module with a domain-specific autoregressive loss (DSAL) they establish a new benchmark to evaluate the efficacy of their method .
Outcome: The proposed method outperforms all baselines and is based on a Sparse Mixture of Experts (SMoE) module .

Similar Papers

HiDe-LLaVA: Hierarchical Decoupling for Continual Instruction Tuning of Multimodal Large Language Model (2025.acl-long)

Copied to clipboard

Challenge: Existing methods to improve instructionfollowing performance of MLLMs often trade off memory efficiency for performance gains, compromising overall efficiency.
Approach: They propose a task-specific expansion and task-general fusion framework based on variations in Centered Kernel Alignment (CKA) similarity across different model layers when trained on diverse datasets.
Outcome: The proposed framework improves performance compared to existing benchmarks.
Tracing Multilingual Knowledge Acquisition Dynamics in Domain Adaptation: A Case Study of Biomedical Adaptation (2026.eacl-long)

Copied to clipboard

Challenge: Multilingual domain adaptation (ML-DA) enables large language models to acquire domain knowledge across languages.
Approach: They propose an adaptive evaluation method that constructs multiple-choice QA datasets from the same bilingual domain corpus used for training.
Outcome: The proposed method constructs multiple-choice QA datasets from the same bilingual domain corpus used for training, thereby enabling direct analysis of multilingual knowledge acquisition.
On Domain-Adaptive Post-Training for Multimodal Large Language Models (2025.findings-emnlp)

Copied to clipboard

Challenge: Adapting general multimodal large language models to specific domains is important for practical applications.
Approach: They investigate domain adaptation of multimodal large language models via post-training . they develop a generate-then-filter pipeline that curates diverse visual instruction tasks .
Outcome: The proposed model outperforms existing models in domain adaptation by combining data from open-source models with training pipelines.
Dynamic Expert Specialization: Towards Catastrophic Forgetting-Free Multi-Domain MoE Adaptation (2025.emnlp-main)

Copied to clipboard

Challenge: Existing approaches to adapt Mixture-of-Experts models to multiple domains are prohibitive computation, cross-domain interference or require separate runs per domain.
Approach: They propose a dynamic expert specialization framework for multi-domain adaptation of Mixture-of-Experts models.
Outcome: The proposed framework reduces forgetting by 89% compared to full fine-tuning as domains scale from 2 to 6 and achieves faster convergence than conventional methods.
PASs-MoE: Mitigating Misaligned Co-drift among Router and Experts via Pathway Activation Subspaces for Continual Learning (2026.acl-long)

Copied to clipboard

Challenge: Existing LoRA-based Mixture-of-Experts (MoE) methods often jointly update the router and experts in an indiscriminate way, causing the router’s preferences to co-drift with experts’ adaptation pathways and exacerbate forgetting.
Approach: They propose a LoRA-induced subspace that reflects which low-rank pathway directions an input activates in each expert, providing a capability-aligned coordinate system for routing and preservation.
Outcome: The proposed method outperforms conventional continual learning baselines and MoE–LoRA variants in accuracy and resistance to forgetting, without increasing model parameters.
GMoE: Global Mixture of Experts with Logit Propagation (2026.acl-long)

Copied to clipboard

Challenge: Sparse Mixture of Experts architectures retain large memory footprints and exhibit significant redundancy, both within and across layers.
Approach: They propose a sparse mixture of experts architecture that uses global experts shared across all layers and adds a Local Expert per layer for layer-specific adaptation.
Outcome: The proposed architecture reduces computational cost by activating only a subset of experts per token while maintaining strong performance.
Enhancing Multimodal Continual Instruction Tuning with BranchLoRA (2025.acl-long)

Copied to clipboard

Challenge: Existing approaches to fine tune Multimodal Large Language Models (MLLMs) are prone to Catastrophic Forgetting (CF) existing approaches rely on the Mixture-of-Experts (MoE) LoRA framework to preserve previous instruction alignments.
Approach: They propose an asymmetric tuning-freezing mechanism to mitigate parameter inefficiency . branch-specific routers are introduced to ensure optimal branch distribution over time .
Outcome: The proposed framework outperforms existing frameworks on the latest MCIT benchmarks.
Soft Orthogonal Low-Rank Adaptation for Knowledge Sharing in Large Language Model Continual Learning (2026.acl-long)

Copied to clipboard

Challenge: Existing methods for continual learning (CL) are designed to mitigate catastrophic forgetting while neglecting knowledge sharing across tasks.
Approach: They propose a framework that facilitates knowledge transfer while mitigating catastrophic forgetting by assigning task-specific parameter subspaces to new tasks . they then leverage attribution scores to evaluate task similarity and employ soft orthogonality between task- specific subspace .
Outcome: The proposed framework facilitates knowledge transfer while mitigating catastrophic forgetting.
Sparser Mixture-of-Adapters with Cross-Layer Generalization (2025.naacl-long)

Copied to clipboard

Challenge: Existing methods for training large language models do not allow sharing adapters across layers . existing methods do not support sharing adapter pools, leading to redundancy and poor generalization .
Approach: They propose a mixture-of-adapter framework that trains a pool of lightweight adapters at each layer and selects the most suitable ones for each input.
Outcome: The proposed framework reduces active adapters by over 85% while boosting task accuracy.
Automatic Expert Discovery in LLM Upcycling via Sparse Interpolated Mixture-of-Experts (2025.acl-long)

Copied to clipboard

Challenge: Sparse Interpolated Mixture-of-Experts (SIMoE) instruction-tuning is an end-to-end algorithm designed to fine-tune a dense pre-trained Large Language Model (LLM) into a MoE-style model that possesses capabilities in multiple specialized domains.
Approach: They propose an algorithm to fine-tune a dense pre-trained Large Language Model into a MoE-style model that possesses capabilities in multiple specialized domains.
Outcome: The proposed algorithm achieves state-of-the-art on common instruction-tuning benchmarks while maintaining an optimal performance-compute trade-off compared to baselines.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations