paper-with-me

홈 › Papers

NEMO: Can Multimodal LLMs Identify Attribute-Modified Objects?

2024-11-26 · Jiaxuan Li, Junwen Mo, MinhDuc Vo, Akihiro Sugimoto, Hideki Nakayama

Multimodal Large Language Models (MLLMs) have made notable advances in visual understanding, yet their abilities to recognize objects modified by specific attributes remain an open question. To address this, we explore MLLMs' reasoning capabilities in object recognition, ranging from commonsense to beyond-commonsense scenarios. We introduce a novel benchmark, NEMO, which comprises 900 images of origiNal fruits and their corresponding attributE-MOdified ones; along with a set of 2,700 questions including open-, multiple-choice-, unsolvable types. We assess 26 recent open-sourced and commercial models using our benchmark. The findings highlight pronounced performance gaps in recognizing objects in NEMO and reveal distinct answer preferences across different models. Although stronger vision encoders improve performance, MLLMs still lag behind standalone vision encoders. Interestingly, scaling up the model size does not consistently yield better outcomes, as deeper analysis reveals that larger LLMs can weaken vision encoders during fine-tuning. These insights shed light on critical limitations in current MLLMs and suggest potential pathways toward developing more versatile and resilient multimodal models.

📄 PDF Abstract BibTeX arXiv:2411.17794

Code (0)

등록된 구현이 없습니다.

Tasks

AttributeMultiple-choiceObject Recognition

Methods 이 논문이 사용한 방법론

SET Dynamic Sparse Training method where weight mask is updated randomly periodically

Similar Papers 제목 키워드 기반

Nemotron 3 Nano Omni: Efficient and Open Multimodal Intelligence

2026-04-27 · NVIDIA, :, Amala Sanjay Deshmukh, Kateryna Chumachenko 외 arxiv

We introduce Nemotron 3 Nano Omni, the latest model in the Nemotron multimodal series and the first to natively support audio inputs alongside text, images, and video. Nemotron 3 Nano Omni delivers consistent accuracy im…

NeMo-Inspector: A Visualization Tool for LLM Generation Analysis

2025-05-01 · Daria Gitman, Igor Gitman, Evelina Bakhturina

Adapting Large Language Models (LLMs) to novel tasks and enhancing their overall capabilities often requires large, high-quality training datasets. Synthetic data, generated at scale, serves a valuable alternative when r…

GSM8KMath

Omni-Embed-Nemotron: A Unified Multimodal Retrieval Model for Text, Image, Audio, and Video

2025-10-03 · Mengyao Xu, Wenfei Zhou, Yauhen Babakhin, Gabriel Moreira 외 arxiv

We present Omni-Embed-Nemotron, a unified multimodal retrieval embedding model developed to handle the increasing complexity of real-world information needs. While Retrieval-Augmented Generation (RAG) has significantly a…

Video Retrieval

NeMo-Aligner: Scalable Toolkit for Efficient Model Alignment

2024-05-02 · Gerald Shen, Zhilin Wang, Olivier Delalleau, Jiaqi Zeng 외

Aligning Large Language Models (LLMs) with human values and preferences is essential for making them helpful and safe. However, building efficient tools to perform alignment can be challenging, especially for the largest…

modelparameter-efficient fine-tuning

A SMART Mnemonic Sounds like "Glue Tonic": Mixing LLMs with Student Feedback to Make Mnemonic Learning Stick

2024-06-21 · Nishant Balepur, Matthew Shu, Alexander Hoyle, Alison Robey 외

Keyword mnemonics are memorable explanations that link new terms to simpler keywords. Prior work generates mnemonics for students, but they do not train models using mnemonics students prefer and aid learning. We build S…

Missing Labels