multimodal large language model
PulseAugur coverage of multimodal large language model — every cluster mentioning multimodal large language model across labs, papers, and developer communities, ranked by signal.
- instance of Gotit.pub 95%
- instance of alphaXiv 90%
- instance of ScienceCast 90%
- instance of CatalyzeX 90%
- instance of Multimodal Large Language Models and Tunings: Vision, Language, Sensors, Audio, and Beyond 90%
- used by CatalyzeX 70%
- instance of DagsHub 70%
- uses text-to-video generation 70%
- used by ScienceCast 60%
- used by Gotit.pub 60%
- used by alphaXiv 60%
- used by DagsHub 50%
16 day(s) with sentiment data
-
New RL method improves AI face anti-spoofing generalization
Researchers have developed a novel reinforcement learning approach called Grounding-Aware Judge Reward Reinforcement Learning (GA-JRRL) to improve the out-of-domain generalization of multimodal large language models (ML…
-
New framework MUSTARD enhances MLLMs for dynamic industrial anomaly detection
Researchers have developed MUSTARD, a novel framework designed to enhance multimodal large language models (MLLMs) for industrial anomaly detection. Unlike previous methods that primarily focus on static visual patterns…
-
New framework ERIM enhances multimodal in-context learning by inducing rules
Researchers have introduced ERIM, a novel framework designed to improve multimodal in-context learning (MICL) by explicitly inducing the underlying rules from demonstration samples. Unlike existing methods that focus on…
-
LLM agents learn from tutorial videos without parameter updates
Researchers have developed Tutorial Agentic Learning (TAL), a novel method enabling Large Language Model (LLM) agents to learn from tutorial videos without requiring parameter updates. This approach consolidates instruc…
-
New Benchmark and Framework Tackle Bias in Long-Video AI Understanding
Researchers have introduced S-MME, a benchmark designed to identify biased behavior in multimodal large language models (MLLMs) when processing long videos. They also propose CLEAR, a training-free framework to address …
-
New IQA Framework Prioritizes Task Utility for Embodied AI
Researchers have proposed a new framework for image quality assessment (IQA) that rethinks the traditional approach by focusing on task utility rather than generic human ratings. This new method, developed using a multi…
-
New research questions foundational premise of neural network pruning
A new paper questions the long-standing foundation of neural network pruning, known as oracle pruning. The research, which involved training and analyzing 37,000 models across various architectures and datasets, found t…
-
PathKGen uses knowledge graphs for explainable histopathology reports
Researchers have developed PathKGen, a novel multimodal large language model (MLLM) designed to improve the explainability of histopathology reports. PathKGen integrates external structured knowledge from pathology-spec…
-
New MulFCoder system generates front-end code for multiple UI frameworks
Researchers have developed MulFCoder, a novel multi-agent system designed to generate front-end code for various UI frameworks like React, Vue, and Angular from UI screenshots. Unlike previous methods that primarily foc…
-
MLLM retrieval heads are sparse, shared across modalities, and optimizable
Researchers have investigated the retrieval heads within multimodal large language models (MLLMs), finding that these heads, responsible for locating relevant information in long inputs, remain sparse and largely shared…
-
Multimodal agent enhances control of scientific instruments with non-stationary responses
Researchers have developed a novel multimodal agent designed to control scientific instruments with non-stationary responses. This agent utilizes three distinct components: deterministic algorithms for extracting symbol…
-
MLLMs automate surgical instrument bounding box annotation, creating large dataset
Researchers have developed a new framework using multimodal large language models (MLLMs) to efficiently generate bounding box annotations for surgical instruments from existing presence labels. This method was tested o…
-
Factor-Aware Predictive Learning enhances cinematic video generation with MLLMs
Researchers have developed Factor-Aware Predictive Learning (FAPL), a novel method to enhance cinematic video generation using multimodal large language models (MLLMs). FAPL trains video diffusion transformers to better…
-
LocalAgent framework enhances MLLM consistency verification
Researchers have introduced LocalAgent, a novel framework designed to enhance the fine-grained instance-level consistency verification capabilities of multimodal large language models (MLLMs). This system employs a coll…
-
New AI systems tackle complex sampling and human-scene interaction
Researchers have developed two new AI systems for complex tasks. One, MASEM, addresses the challenge of sampling from disconnected probability distributions, improving efficiency in areas like Bayesian optimization and …
-
DRAPE framework enhances multimodal LLMs with dynamic prompt generation
Researchers have developed DRAPE, a novel prompt-learning framework designed for Multimodal Continual Instruction Tuning (MCIT). This method generates instance-specific soft prompts by conditioning on both textual instr…
-
New ABACUS benchmark reveals image editing models struggle with object counting
Researchers have introduced ABACUS, a new benchmark and evaluation protocol designed to accurately assess the numerical capabilities of image editing models. Current methods often overlook errors in object counting and …
-
New framework AnomalyJev evaluates MLLM anomaly detection explanations
Researchers have developed AnomalyJev, a new framework for evaluating the explanations generated by multimodal large language models (MLLMs) in anomaly detection tasks. AnomalyJev extracts verifiable claims from explana…
-
New TV-AudioRemover framework enables multimodal sound removal from videos
Researchers have introduced TV-AudioRemover, a novel framework designed to remove sounds associated with specific visual objects from videos. Unlike previous methods that relied on single-modal control, TV-AudioRemover …
-
CERTH-ITI details MLLM-based approaches for TRECVID 2025 video search and QA
Researchers from CERTH-ITI have detailed their approach for the TRECVID 2025 Ad-hoc Video Search and Video Question Answering tracks. For ad-hoc video search, they employed a two-stage framework using multiple vision-la…