Unverified50% confidenceConceptExact time
MMEB (Massive Multimodal Embedding Benchmark) is a unified benchmark for evaluating general-purpose multimodal embedding models, covering classification, retrieval, visual question answering, grounding and many other tasks and numerous datasets, emphasizing horizontal comparison of different strategies under the same data and training conditions.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedMMLU (Massive Multitask Language Understanding) and HumanEval are representative general benchmarks, used for commonsense reasoning and code generation evaluation respectively.70% similarUnverifiedW&B (Weights & Biases) is a mainstream experiment tracking platform in machine learning engineering, used to record metadata such as hyperparameters, metric curves, and system resource usage for each training job.69% similarUnverifiedTable Batched Embedding (TBE) is a fundamental operator in recommendation model training, responsible for mapping sparse ID features to dense vectors and supporting batched queries across multiple GPUs and multiple embedding tables.68% similarUnverifiedBenchmarks such as MMVP and MMStar are specifically designed to detect the degree of visual dependency, deliberately including questions that cannot be answered correctly through language priors alone.66% similarUnverifiedBEIR Benchmark是一套跨领域的检索评估基准，包含多个数据集和标准指标（nDCG、Recall），是系统对比检索策略的首选。65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1028181API
curl https://kongchang.com/api/v1/knowledge/claims/1028181MCP
get_claim(id=1028181)