跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 8 天前AI 評分43

LogT:透過圖像進行邏輯推理以支援視覺搜尋

LogT: Logically Think with Images for Visual Search

AI 導讀

LogT 提出全自動 SFT 數據引擎,透過邏輯鏈引導推理,生成具挑戰性的視覺問題及結構化工具使用軌跡。以 Qwen2.5-VL-7B 為基礎的 LogT-SFT 使用的數據少 3.1 倍、GPU 時數少逾 120 倍,表現仍超越結合 SFT 和 RL 訓練的 SOTA 方法。只採用簡單 RL 配方且不作工具專用設計,模型表現亦較先前複雜 RL 設定高 9.59 分。

正文

Published on Nov 14

Authors:

,

,

,

,

,

,

,

Abstract

Thinking with images empowers vision-language models (VLMs) to use an external Zoom-in tool for fine-grained visual search. Recent works employ supervised fine-tuning (SFT) and reinforcement learning (RL) with tool-specific designs to enable such capabilities. However, achieving logical reasoning alongside accurate and efficient tool use remains a significant challenge. To address this, we shift the focus away from increasing RL complexity and instead identify SFT data quality as a key bottleneck. We then propose LogT (short for Logically Think with Images), a fully automated SFT data engine that explicitly enforces logic-chain-guided reasoning. LogT generates an SFT dataset featuring challenging, visually grounded questions and structured tool-use traces with human-like logic. We validate our approach by fine-tuning two base models, Qwen2.5-VL-7B and InternVL3.5-8B, on the dataset and evaluating them across seven benchmark test sets. We show that LogT-SFT based on Qwen2.5-VL-7B outperforms the state-of-the-art trained with both SFT and RL, despite using 3.1times less data and >120times fewer GPU hours. Furthermore, applying a simple RL recipe without tool-specific design on LogT-SFT yields a model that surpasses prior methods with complex RL setups by a substantial 9.59-point margin. In addition to strong performance gains, our model exhibits higher tool-use accuracy and efficiency, as well as more structured and logically consistent reasoning traces, directly validating the effectiveness of our approach.

View arXiv page View PDF Add to collection

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2611.19200 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2611.19200 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2611.19200 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co