LeapLabTHU/Simple-WAM

Official implementation of "What Makes World Action Models Generalize? An Empirical Study of Test-Time Future Modeling"

Science & researchPython

0.4 Cooling Magnitude out of 10 — how fast interest is growing, not a quality score. Star growth looks organic. Data as of October 2, 2026, 01:39 UTC.

Open on Seismograph Open on GitHub

About the project

Official implementation of a study on what makes world action models generalize, with an empirical study of test-time future modeling. Includes training and evaluation code on LIBERO and RoboTwin benchmarks.

Useful for

Evaluate checkpoints on the LIBERO benchmark using run_libero_manager.py
Run task-generalization evaluation on held-out LIBERO suites with taskgen suffixes
Test model robustness to image corruptions on LIBERO-plus

README summarized by DeepSeek V4.1 Flash. Details may be inaccurate.

Why it’s trending

The repository is 4 days old and already has 62 stars. With less than two weeks of history, there's no usual pace to compare the spike against yet.
Top new repositories this week: #186.

Stars per day

Bars are daily stars, the line is the usual pace. Red marks spike days.

Numbers

Total stars: 62
Today: 0 · ≈ 1 by evening
Forks: 1
Issues and pull requests: 0
Watchers: 1
Language: Python
Created: September 28, 2026
Last push: September 30, 2026

Star trust

Growth looks organic: forks and discussion are in line with active projects, and stars arrive unevenly, the way people give them.

These are heuristics, not a verdict: we judge by the repository’s behavior, not by a list of stargazers.

Spotted in

October 2, 2026Top new repositories this week: #183

README badge

Paste it into your README — the badge shows the current magnitude and links to this page.

[![Seismograph](https://gitnova.dev/badge/LeapLabTHU/Simple-WAM.svg?lang=en)](https://gitnova.dev/en/r/LeapLabTHU/Simple-WAM)

More in this category

3.1
latticeverse/LatticeVerse
An umbrella research map and integration layer for lattice metamaterials: links papers and implementations for geometry modeling, physics simulation, inverse design, and manufacturing-aware optimization.

Early signalScience & research+0 stars today, ≈ 30 by evening
2.5
OpSafari/hypoarena
Fully offline workbench for testing components of scientific hypothesis-discovery pipelines: hypothesis/evidence graphs, synthetic literature with planted causal chains, paraphrase dedup, Elo tournaments, and Bayesian…

Early signalScience & researchPython+0 stars today, ≈ 6 by evening
2.3
apromisedland/Tame_3D
Python research framework for multi-agent 3D situated reasoning: a manager plans typed geometric programs, egocentric and global experts read local views and geometry, a critic checks evidence, and conformal…

Early signalScience & researchPython+0 stars today, ≈ 21 by evening
1.8
apromisedland/VLCoT
Research implementation of VLCoT: stage-verifiable latent chain-of-thought reasoning with observation-grounded verification and dependency-aware local repair for long-horizon robotic manipulation. Includes an…

Early signalScience & researchPython+0 stars today, ≈ 13 by evening
1.6
aipoch/open-science
Local-first desktop workbench for scientific research with AI agents: runs Python/R, connects scientific data sources, and tracks artifact provenance for reproducibility.

CoolingScience & researchTypeScript+0 stars today, ≈ 11 by evening
1.3
amap-cvlab/ABot-Recon
ABot-Recon is a streaming 3D reconstruction model that uses a fixed 12-frame local context per step, predicting a point map and adjacent relative pose, then composing them into a global trajectory and point cloud.

SteadyScience & researchPython+0 stars today, ≈ 9 by evening