TGLG Benchmark

What is TGLG?

Temporally-Grounded Language Generation (TGLG) is a benchmark for real-time vision-language models (VLMs) that focus on two key capabilities: perceptual…

Released

2025-05-09

Modality

multimodal

Evaluates

多模态, Multimodal, 语言, Language, Strong Reasoning, VQA, 多模态模型, VLM, 语言理解, Comprehension, 逻辑推理, Reasoning, 不支持, Unsupported

Openness

open

Importer

OpenCompass Hub

Reported scores

0

Source provenance

• Original evidence https://hub.opencompass.org.cn/dataset-detail/TGLG

TGLG paper, code and dataset

• Paper https://arxiv.org/pdf/2505.11326
• Code https://github.com/yukw777/tglg
• Dataset https://huggingface.co/datasets/kpyu/tglg
• Dataset https://huggingface.co/datasets/kpyu/tglg.

Which sources cite TGLG?

• TGLG

Related 多模态, Multimodal, 语言, Language, Strong Reasoning, VQA, 多模态模型, VLM, 语言理解, Comprehension, 逻辑推理, Reasoning, 不支持, Unsupported benchmarks

• 3DCodeBench
• A-Bench
• A-OKVQA
• AbsPyramid
• ActionAtlas
• ActiveFly-Bench
• Ada-LEval
• AdvancedMathBench