Ambient @ EgoLongQA 2026: Distilling Long-Video perception into a Sub-2B Model
- 类型：arxiv
- 标识：2609.07154
- 链接：https://arxiv.org/abs/2609.07154
- 主分类：multimodal
- 形态：method
- 被引：0
- 被引来源：Semantic Scholar + OpenAlex
- S2被引：0
- OpenAlex被引：0
- 影响力被引：0
- TLDR：The system is a single 2B vision-language model that answers multiple-choice questions about ten-minute egocentric videos in one greedy forward pass, and reaches 89% of the accuracy of the large agentic pipeline using 1.1% of its parameters.
- OpenAlex ID：W7212037653
- OpenAlex DOI：10.48550/arxiv.2609.07154
- DOI：10.48550/arxiv.2609.07154
- DOI来源：OpenAlex
- 开放获取：green
- 开放获取链接：https://arxiv.org/pdf/2609.07154
- OpenAlex更新：2026-09-16
- 副分类：agent
- 待LLM分类：否
- 标题中文：Ambient @ EgoLongQA 2026: Distilling Long-Video perception into a Sub-2B Model
- TLDR中文：该系统仅用一个 2B 视觉-语言模型，通过一次贪心前向传播即可回答关于十分钟第一人称视频的多选题，并仅用大型 Agentic pipeline 1.1% 的参数即达到其 89% 的准确率。
- 来源文件：
- /inbox/tom/_candidates/2026-09-15-agent-rag-longcontext-candidates.json
- [OpenAlex backfill]
- [S2 enrich]