Datasetpublic

AgentVidBench123/agentvidbench

AgentVidBench: A Multi-Hop Video Question Answering Benchmark for Evaluating MLLM Agents Agentic Video Understanding Benchmark — 100 multiple-choice video QA questions, 26 options each (A-Z; ~3.8% random baseline) Anonymous authors — under review. Layout . ├── README.md ├── questions.jsonl # 100 rows — one per question ├── videos.jsonl # 71 rows — one per unique video ├── videos/ │ └── video*.mp4 # 71 video files └── transcripts/… See the full description on the dataset page: https://huggingface.co/datasets/AgentVidBench123/agentvidbench.

sourceHugging Faceccupdated 13d agoView on Hugging Face

0likes67downloads

Dataset Card

AgentVidBench: A Multi-Hop Video Question Answering Benchmark for Evaluating MLLM Agents

Agentic Video Understanding Benchmark — 100 multiple-choice video QA questions, 26 options each (A-Z; ~3.8% random baseline)

Anonymous authors — under review.

Layout

.
├── README.md
├── questions.jsonl        # 100 rows — one per question
├── videos.jsonl           # 71 rows — one per unique video
├── videos/
│   └── video*.mp4         # 71 video files
└── transcripts/
    └── video*.srt         # 71 Whisper transcripts (28 are empty for silent videos)

`questions.jsonl` schema

Field	Type	Notes
`question_id`	int	1–100
`video_path`	str	relative path to the video file (`videos/videoN.mp4`); equals the `file_name` in `videos.jsonl`
`transcript_path`	str	relative path to the Whisper transcript (`transcripts/videoN.srt`); shares the basename with `video_path`
`title`	str	short descriptive title
`question_text`	str	full question prompt
`options`	list[str] (len 26)	answer choices in order A, B, C, … Z
`answer`	str	correct answer letter, A–Z
`answer_explanation`	str	reasoning + reference for the answer
`skills`	list[str]	core capabilities the question tests
`difficulty`	str	`easy` / `medium` / `hard` / `very_hard`
`options_type`	str	distractor strategy
`categories`	list[str]	cognitive task tags
`milestones`	list[step]	sub-tasks the question tests; each step is `{id, type, description}`
`ecom_required`	bool	true iff the answer cannot be locked in until later evidence in the video has been observed (i.e., the full clip must be watched)
`trajectory`	dict	Reasoning trace. Shape: `{steps: list[step], final_answer: str, model: str, temperature: float, generated_at: str, elapsed_seconds: float}`. Each step is `{tool: str, args: dict, thought: str, observation: str}`; `args` is heterogeneous per tool (HF infers it as a `Json` feature). Populated for all 100 questions.

`videos.jsonl` schema

Field	Type	Notes
`file_name`	str	relative path to the video file (`videos/videoN.mp4`)
`title`	str	—
`duration`	str	`M:SS`
`genre`	str	one of 12 enum values
`production_style`	list[str]	shot/edit style tags
`audio_type`	list[str]	audio characteristics
`length_category`	str	`short` / `medium` / `long`
`url`	str	source URL
`license`	str	license
`source`	str	—
`author`	str	—

Download

bash

hf download AgentVidBench123/agentvidbench --repo-type dataset --local-dir dataset

Citation

@misc{anonymous2026agentvidbench,
  title  = {AgentVidBench: A Multi-Hop Video Question Answering Benchmark for Evaluating MLLM Agents},
  author = {Anonymous},
  year   = {2026},
  note   = {Under review}
}