[NeurIPS 2025 Spotlight] Towards Safety Alignment of Vision-Language-Action Model via Constrained Learning.
active 2025-02-27 → 2026-04-26 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 424 days from 2025-02-27 to 2026-04-26. Pushes: 8 total, peak 2 in a day. Pull requests: 2 total, peak 1 in a day. Issues: 19 total, peak 2 in a day. Comments: 15 total, peak 2 in a day. Stars: 77 total, peak 5 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| Ethyn13 | 18 | 5 | 2 | 6 |
| muchvo | 9 | 3 | 0 | 4 |
| rdesc | 4 | 0 | 0 | 2 |
| insightHalloween | 3 | 0 | 0 | 1 |
| yangyan010119 | 2 | 0 | 0 | 1 |
| y-1-2 | 1 | 0 | 0 | 0 |
| fujiawei0724 | 1 | 0 | 0 | 0 |
| tayalmanan28 | 1 | 0 | 0 | 0 |
| heartInsert | 1 | 0 | 0 | 0 |
| Galaxyphan | 1 | 0 | 0 | 0 |
| NielsRogge | 1 | 0 | 0 | 1 |
| BaoshunWq | 1 | 0 | 0 | 0 |
| autodriving-heart | 1 | 0 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
- Issue#26muchvo2026-03-14 08:12Why is this line commented out ?
- Issue#25insightHalloween2026-02-18 08:36SafeVLA的视觉编码器
- Issue comment#25insightHalloween2026-02-18 08:36SafeVLA的视觉编码器
- Issue#25insightHalloween2026-02-07 09:22SafeVLA的视觉编码器
- Issue comment#23rdesc2026-01-21 05:34OOM error when running training script
- Issue comment#24yangyan0101192026-01-16 03:11SafePPOLogGrad implement
- Issue comment#24Ethyn132026-01-16 03:06SafePPOLogGrad implement
- Issue#24yangyan0101192026-01-16 02:46SafePPOLogGrad implement
- Issue#22rdesc2026-01-13 19:03how long does training take?
- Issue#22rdesc2026-01-12 21:23how long does training take?
- Issue comment#20rdesc2026-01-12 17:40Typo in readme
- Pull request#21Ethyn132026-01-11 08:45
- Pull request#17Ethyn132025-12-04 07:39
- Issue comment#16Ethyn132025-12-04 07:29THOR object not found
- Issue#15Ethyn132025-12-02 04:13The download_il_ckpt.py file don't exist
- Issue#14tayalmanan282025-11-25 10:04Add Instructions for Running SafeVLA Without Docker
- Issue comment#12Ethyn132025-11-11 07:47The environment of safety-gymnasium lacks some important attribute about the AbstractSPOCTask class
- Issue#12BaoshunWq2025-11-10 02:12The environment of safety-gymnasium lacks some important attribute about the AbstractSPOCTask class
- Issue#6Ethyn132025-10-31 06:46Missing .hdf5 training data files
- Issue#9Ethyn132025-09-05 14:56缺少'SafePPOValue'
- Issue comment#9Ethyn132025-09-04 19:26缺少'SafePPOValue'
- Issue#9y-1-22025-09-04 09:16缺少'SafePPOValue'
- Issue comment#4Ethyn132025-08-10 06:34Downloading assets and annotations command
- Issue comment#8Ethyn132025-08-10 06:34Imitation learning data and imitation learning launch script
- Issue#7fujiawei07242025-07-27 01:13Launch safe vla environment using safe gymnasium
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 77 stars here means stars gained during the window, not the repo's star count.