The hub for EleutherAI's work on interpretability and learning dynamics
active 2023-08-15 → 2026-07-02 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 1053 days from 2023-08-15 to 2026-07-02. Pushes: 81 total, peak 9 in a day. Pull requests: 54 total, peak 4 in a day. Issues: 69 total, peak 11 in a day. Comments: 109 total, peak 7 in a day. Stars: 1,212 total, peak 16 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 |
|---|---|---|---|---|
| StellaAthena | 65 | 29 | 6 | 16 |
| haileyschoelkopf | 57 | 18 | 10 | 12 |
| Quentin-Anthony | 28 | 9 | 10 | 2 |
| uSaiPrashanth | 26 | 15 | 4 | 5 |
| CLAassistant | 15 | 0 | 0 | 15 |
| oskarvanderwal | 12 | 9 | 1 | 2 |
| itsnamgyu | 9 | 0 | 0 | 7 |
| pietrolesci | 9 | 0 | 1 | 6 |
| yuzc19 | 7 | 0 | 0 | 6 |
| sunnyddelight | 6 | 0 | 6 | 0 |
| segyges | 5 | 0 | 4 | 0 |
| markschoene | 4 | 0 | 0 | 3 |
| efittschen | 4 | 0 | 1 | 2 |
| prakharg24 | 4 | 0 | 0 | 1 |
| javirandor | 4 | 0 | 0 | 1 |
| Lisennlp | 3 | 0 | 0 | 1 |
| activatedgeek | 3 | 0 | 0 | 1 |
| borgr | 3 | 0 | 1 | 0 |
| ParthaKrPaul | 3 | 0 | 0 | 1 |
| mirandrom | 2 | 0 | 0 | 1 |
Recent activity
Latest issues, pull requests and releases
- Issue#197impacterai2026-04-04 00:05Add Repurpose AI — AI content repurposing tool
- Issue comment#195timethink2026-03-19 03:04Pythia-2.8B PT CheckPoints is same between steps 0 to steps 6300
- Issue comment#195snizio2026-03-13 18:34Pythia-2.8B PT CheckPoints is same between steps 0 to steps 6300
- Issue comment#123philipzeiler2025-08-26 13:33Batch Viewer : Why Sequence Length 2049?
- Issue comment#159Deven141252025-07-30 14:56Convert to GGUF
- Issue#190efittschen2025-07-24 21:14seed2 shuffle_idx.npy corrupted
- Issue comment#189Quentin-Anthony2025-06-09 18:33print fix in example
- Issue comment#189CLAassistant2025-06-09 03:49print fix in example
- Pull request#189ashwanirathee2025-06-09 03:49
- Pull request#188Quentin-Anthony2025-06-02 17:24
- Pull request#188rdiehlmartinez2025-06-02 16:34
- Issue comment#167StellaAthena2025-05-23 11:42open-source the training data used between two adjacent checkpoints
- Issue#167StellaAthena2025-05-23 11:41open-source the training data used between two adjacent checkpoints
- Issue#121StellaAthena2025-05-23 11:41Any results on the validation set?
- Issue#144StellaAthena2025-05-23 11:40[Pythia on Pile-Dedup] Training for ~1.5 epochs: how to identify the repeated sequences (i.e., the additional .5 epoch)?
- Issue comment#144StellaAthena2025-05-23 11:40[Pythia on Pile-Dedup] Training for ~1.5 epochs: how to identify the repeated sequences (i.e., the additional .5 epoch)?
- Issue#135StellaAthena2025-05-23 11:39Inconsistent init methods of pythia-6.9b model
- Issue comment#149StellaAthena2025-05-23 11:35"gas" configuration doesn't do anything
- Issue#149StellaAthena2025-05-23 11:35"gas" configuration doesn't do anything
- Issue#162StellaAthena2025-05-23 11:30Pythia 12b flash config
- Issue#160StellaAthena2025-05-23 11:30how to use Pythia
- Issue comment#127StellaAthena2025-05-23 11:26Has the data been shuffled?
- Issue#127StellaAthena2025-05-23 11:26Has the data been shuffled?
- Issue#165StellaAthena2025-05-23 11:25Issue while showering NLO events with NLO
- Issue comment#175StellaAthena2025-05-23 11:19No EOD Tokens in EleutherAI/pile-deduped-pythia-preshuffled
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 1,212 stars here means stars gained during the window, not the repo's star count.