Search results

Predictor mask token mismatch

Status: Open.

#124 In facebookresearch/vjepa2;

· mirlanium opened on Feb 23, 2026

How can I use multi-layer feature maps from the V-JEPA model to perform classification tasks such as ImageNet? Does the repo provide any example code for this? I couldn’t find it

Status: Open.

#122 In facebookresearch/vjepa2;

· zhxie0117 opened on Feb 17, 2026

Using VJEPA2 Frozen Encoder as Video Feature Extractor

Status: Open.

#120 In facebookresearch/vjepa2;

· norman-nhan opened on Feb 6, 2026

tubelet_size and frameskip

Status: Open.

#118 In facebookresearch/vjepa2;

· Cc1020-hh opened on Feb 3, 2026

Temporal features from frames

Status: Open.

#117 In facebookresearch/vjepa2;

· zillur01 opened on Feb 1, 2026

Add Lightweight Backbones to Enable Applied & Low‑Resource Research

Status: Open.

#112 In facebookresearch/vjepa2;

· VasylVasylenko opened on Jan 2, 2026

如何将我的后训练vjepa2部署到我自己的机械臂（ac版本的模型也是我基于我的数据集构造的）

Status: Open.

#111 In facebookresearch/vjepa2;

· kaileliu opened on Dec 22, 2025

AC model not working in simulation

Status: Open.

#110 In facebookresearch/vjepa2;

· ProxyCausal opened on Dec 19, 2025

Transforms for the QA Dataset

Status: Open.

#108 In facebookresearch/vjepa2;

· andyshuo opened on Dec 6, 2025

Stereo videos in DROID raw: included in V-JEPA 2-AC training?

Status: Open.

#107 In facebookresearch/vjepa2;

· nisarganc opened on Dec 4, 2025

Bug in clip resizing

Status: Open.

#106 In facebookresearch/vjepa2;

· aderylo opened on Dec 4, 2025

Processing image data for V-JEPA2 for QA

Status: Open.

#105 In facebookresearch/vjepa2;

· andyshuo opened on Dec 4, 2025