Research map / Vision-language and multimodal / Robot action and navigation
ADAPT: Vision-Language Navigation With Modality-Aligned Action Prompts
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add modality-aligned action prompts | VLN↻BERT with ResNet-152 · R2R Val Unseen | SR: 63 -> 64 | Helped |
| add modality-aligned action prompts | VLN↻BERT with ResNet-152 · RxR Val Unseen | SR: 45.5 -> 46.7 | Helped |
| add modality-aligned action prompts (ablation) | VLN↻BERT with CLIP features · R2R Val Unseen | SR: 61.5 -> 63.1 | Helped |
| add modality alignment loss to action prompts (ablation) | ADAPT-1 · R2R Val Unseen | SR: 61.6 -> 62.8 | Helped |
| add sequential consistency loss to action prompts with modality alignment loss (ablation) | ADAPT-2 · R2R Val Unseen | SR: 62.8 -> 63.1 | Helped |
| replace action-level prompts with object words and related object images (ablation) | ADAPT · R2R Val Unseen | SR: 61.9 vs 62.5 | Hurt |
| use action prompts as training samples directly instead of explicit prompt learning (ablation) | ADAPT · R2R Val Unseen | SR: 59.0 vs 62.5 | Hurt |
| select image sub-prompts from whole paths rather than Fine-grained R2R sub-paths (ablation) | ADAPT · R2R Val Unseen | SR: 61.8 vs 62.5 | Hurt |
| use CLIP visual features instead of ResNet-152 features | VLN↻BERT and ADAPT · R2R Val Unseen | SR: 63 -> 65; 64 -> 66 | Mixed |
| train with ADAPT instead of VLN↻BERT | VLN↻BERT · R2R Val Unseen, Scan 60% | SR: 57.7 -> 57.2 | Hurt |
Closest papers by meaning
- Vision-Language Models Create Cross-Modal Task Representations (ICML 2025 · similarity 0.89)
- ATI-VLA: Action-Centric Predictive Vision–Language–Action Models via Actionable Alignment Then Adaptive Injection (NeurIPS 2026 · similarity 0.88)
- CogVLA: Cognition-Aligned Vision-Language-Action Models via Instruction-Driven Routing & Sparsification (NeurIPS 2025 · similarity 0.88)
- Not All Features Are Created Equal: A Mechanistic Study of Vision-Language-Action Models (NeurIPS 2026 · similarity 0.88)
- FineVLA: Fine-Grained Instruction Alignment for Steerable Vision-Language-Action Policies (NeurIPS 2026 · similarity 0.88)
- Human-Aware Vision-and-Language Navigation: Bridging Simulation to Reality with Dynamic Human Interactions (NeurIPS 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.