Pith. sign in

REVIEW 6 cited by

Beyond Anti-Forgetting: Multimodal Continual Instruction Tuning with Positive Forward Transfer

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 2401.09181 v3 pith:NSRQRM3C submitted 2024-01-17 cs.LG

classification cs.LG
keywords forwardtransfertaskstuningcatastrophicforgettinginstructionmcit
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Multimodal Continual Instruction Tuning (MCIT) enables Multimodal Large Language Models (MLLMs) to meet continuously emerging requirements without expensive retraining. MCIT faces two major obstacles: catastrophic forgetting (where old knowledge is forgotten) and negative forward transfer (where the performance of future tasks is degraded). Although existing methods have greatly alleviated catastrophic forgetting, they still suffer from negative forward transfer. We discover a large discrepancy in different input embeddings by performing singular value decomposition (SVD) on input embeddings. This discrepancy results in the model learning irrelevant information for old and pre-trained tasks, leading to catastrophic forgetting and negative forward transfer. To address these issues, we propose Prompt Tuning with Positive Forward Transfer (Fwd-Prompt), a prompt-based method that projects the prompt gradient to the residual space to minimize interference between tasks and to the pre-trained subspace for reusing pre-trained knowledge. Our experiments demonstrate that Fwd-Prompt achieves state-of-the-art performance while updating fewer parameters and requiring no old samples. Our research illuminates the potential of continuously adapting MLLMs to new tasks under the instruction tuning paradigm and encourages future studies to explore MCIT.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

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

  1. Domain-Aware Scaling Laws Uncover Data Synergy

    cs.LG 2026-07 conditional novelty 7.0 of 10

    Domain-aware scaling laws with fitted γ and σ synergy terms recover stable code-math interactions from observational LLM mixtures and correctly predict mixture rankings in controlled small-scale trainings.

  2. Harmonious Parameter Adaptation in Continual Visual Instruction Tuning for Safety-Aligned MLLMs

    cs.CV 2025-11 conditional novelty 6.0 of 10

    HPA is a post-training parameter-selection method that keeps safety-aligned multimodal LLMs safe and reduces forgetting during continual visual instruction tuning.

  3. Dynamic Mixture of Curriculum LoRA Experts for Continual Multimodal Instruction Tuning

    cs.CV 2025-06 conditional novelty 6.0 of 10

    D-MoLE dynamically allocates LoRA experts to the most gradient-sensitive layers of a vision-language model for each new task, and routes inputs via per-task autoencoders, reporting near-zero forgetting on a nine-task ...

  4. Continual Gradient Low-Rank Projection Fine-Tuning for LLMs

    cs.LG 2025-07 conditional novelty 5.0 of 10

    GORP jointly trains LoRA and full-rank parameters inside a low-rank gradient subspace built from Adam first moments, reporting higher average accuracy and lower forgetting than O-LoRA and N-LoRA on LLM continual learn...

  5. Enhancing Multimodal Continual Instruction Tuning with BranchLoRA

    cs.CL 2025-05 conditional novelty 5.0 of 10

    BranchLoRA reduces catastrophic forgetting in multimodal continual instruction tuning by using a shared LoRA matrix, task-specific branches, frozen experts, and learned task keys.

  6. Continual Learning for Generative AI: From LLMs to MLLMs and Beyond

    cs.LG 2025-06 conditional novelty 4.0 of 10

    A survey that categorizes continual learning methods for generative models into architecture-based, regularization-based, and replay-based paradigms across four model families.

Pith tools