Li et al.: querying-transformer bridge between frozen vision and language models

What changed

Trained a lightweight querying transformer in representation-learning and generative-alignment stages to bridge frozen pretrained image encoders and language models.

Scientific scope

Trained a lightweight querying transformer in representation-learning and generative-alignment stages to bridge frozen pretrained image encoders and language models. Frozen unimodal models and prior multimodal bridges are inherited. Fewer trainable parameters does not mean the pretrained backbones or their original training cost disappear. Zero-shot instruction examples do not prove general image understanding.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Frozen unimodal models and prior multimodal bridges are inherited. Fewer trainable parameters does not mean the pretrained backbones or their original training cost disappear. Zero-shot instruction examples do not prove general image understanding.

Contributors

• Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.

• BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
• Li et al.: querying-transformer bridge between frozen vision and language models — original project

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction