読書リサーチレーダー投資フレームワーク
ログイン / 新規登録
← 読解に戻る
ニュースクイックテイク1 分で読めるDisconfirmAI 編集部

Small Model Cuts Video Agent Costs

New training framework lets a 4B model match Gemini on video benchmarks while slashing API token usage.

2026-10-06 12:00 GMT+8

The VideoResearchAgent framework enables Qwen3.5-4B to achieve 40.48% accuracy on the Video-BrowseComp benchmark, comparable to Gemini-3-Flash-Preview.

This result comes from a preprint submitted on October 4. Traditional video agents rely on slow, unreliable live web interactions, while fixed local simulations often induce retrieval-specific shortcuts that fail to generalize to the open web.

The team introduced the RDR-GRPO algorithm, which randomizes candidate rankings, distractors, and metadata during training to reduce overfitting. Compared to the untrained model, cumulative API token consumption dropped by 74.9%.

These are author-reported results based on a specific simulator environment; independent reproduction on real open-web video has not yet been verified.

ソース

  • arxiv.org↗
リサーチを続ける →
ログイン / 新規登録
読書リサーチレーダー投資フレームワーク