AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

AstraNav-Memory: Contexts Compression for Long Memory

Junjun Hu, Xinda Xue, Botao Ren, Minghua Luo, Jintao Chen, Haochen Bai, Liangliang You, Mu Xu

CVPR 2026 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Vision-language navigation agents · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (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 changedWhereMeasuredEffect
compress visual tokens to about 30 per image with DINOv3 + PixelUnshuffle and keep long image historyQwen2.5-VL-3B navigation policy · GOAT-Bench Val-UnseenSR: 47.2 -> 62.7 vs MTU3DImproved
sameQwen2.5-VL-3B · HM3D-OVON Val-UnseenSR: 40.8 -> 62.5 vs MTU3D*Improved
apply 16x token compression to 50 stored images (ablation)navigation policy · GOAT subsettrain time per iter: 26.4 s -> 6.5 sImproved
apply 16x token compression to 50 stored images (ablation)navigation policy · GOAT subsetGPU memory: 90.6 GB -> 46.8 GBImproved
apply 16x token compression to 50 stored images (ablation)navigation policy · GOAT subsetSR: 60.2 -> 56.6Hurt
increase stored images from 50 to 100 at 16x compression (ablation)navigation policy · GOAT subsetSR: 56.6 -> 57.5Improved
increase stored images from 100 to 200 at 16x compression (ablation)navigation policy · GOAT subsetSR: 57.5 -> 55.2Hurt
use 4x compression with 100 images instead of 1x with 50 (ablation)navigation policy · GOAT subsetSR: 60.2 -> 63.2Improved
use 64x compression instead of 16x (ablation)navigation policy · GOAT subsetSR: 56.6 -> 42.5 (50 images)Hurt
use DINOv3 ViT-M instead of ViT-S (ablation)tokenizer backbone · GOAT subsetSR: 53.3 -> 57.5Improved
use DINOv3 ViT-L instead of ViT-M (ablation)tokenizer backbone · GOAT subsetSR: 57.5 -> 59.4Improved
use DINOv3-based compressed encoder instead of original Qwen ViT (ablation)navigation policy · per-category SR (Table 5)SR: e.g. freezer 72 -> 88; carpet 61 -> 56Mixed

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy