Correlating instruction-tuning (in multimodal models) with vision-language processing (in the brain)
Transformer-based language models, though not explicitly trained to mimic brain recordings, have demonstrated surprising alignment with brain activity. Progress in these models-through increased size, instruction-tuning, and multimodality-has led to better representational alignment with neural data. Recently, a new class of instruction-tuned multimodal LLMs (MLLMs) have emerged, showing remarkable zero-shot capabilities in open-ended multimodal vision tasks. However, it is unknown whether MLLMs, when prompted with natural instructions, lead to better brain alignment and effectively capture instruction-specific representations. To address this, we first investigate brain alignment, i.e., measuring the degree of predictivity of neural visual activity using text output response embeddings from MLLMs as participants engage in watching natural scenes. Experiments with 10 different instructions show that MLLMs exhibit significantly better brain alignment than vision-only models and perform comparably to non-instruction-tuned multimodal models like CLIP. We also find that while these MLLMs are effective at generating high-quality responses suitable to the task-specific instructions, not all instructions are relevant for brain alignment. Further, by varying instructions, we make the MLLMs encode instruction-specific visual concepts related to the input image. This analysis shows that MLLMs effectively capture count-related and recognition-related concepts, demonstrating strong alignment with brain activity. Notably, the majority of the explained variance of the brain encoding models is shared between MLLM embeddings of image captioning and other instructions. These results suggest that enhancing MLLMs' ability to capture task-specific information could lead to better differentiation between various types of instructions, and thereby improving their precision in predicting brain responses.
Code (1)
Tasks
Image CaptioningMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
MLAN: Language-Based Instruction Tuning Improves Zero-Shot Generalization of Multimodal Large Language Models
We present a novel instruction tuning recipe to improve the zero-shot task generalization of multimodal large language models. In contrast to existing instruction tuning mechanisms that heavily rely on visual instruction…
Instruction FollowingZero-shot GeneralizationVisual Instruction Tuning
Instruction tuning large language models (LLMs) using machine-generated instruction-following data has improved zero-shot capabilities on new tasks, but the idea is less explored in the multimodal field. In this paper, w…
1 Image, 2*2 Stitching3D Question Answering (3D-QA)Image ClassificationImage Retrieval+9InstructionGPT-4: A 200-Instruction Paradigm for Fine-Tuning MiniGPT-4
Multimodal large language models are typically trained in two stages: first pre-training on image-text pairs, and then fine-tuning using supervised vision-language instruction data. Recent studies have shown that large l…
Instruction FollowingQuestion AnsweringVisual Question AnsweringTraining-Free Mitigation of Language Reasoning Degradation After Multimodal Instruction Tuning
Multimodal models typically combine a powerful large language model (LLM) with a vision encoder and are then trained on multimodal data via instruction tuning. While this process adapts LLMs to multimodal settings, it re…
GSM8KLanguage ModelingLanguage ModellingLarge Language Model+1Position-Enhanced Visual Instruction Tuning for Multimodal Large Language Models
Recently, Multimodal Large Language Models (MLLMs) that enable Large Language Models (LLMs) to interpret images through visual instruction tuning have achieved significant success. However, existing visual instruction tu…
cross-modal alignmentPosition