Extension of TransformerLens to different Transformer architecture refinements (DiffAttention, etc.)
Python · active 2025-02-15 → 2026-08-07 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 539 days from 2025-02-15 to 2026-08-07. Pushes: 1,251 total, peak 64 in a day. Pull requests: 244 total, peak 15 in a day. Issues: 62 total, peak 12 in a day. Comments: 132 total, peak 12 in a day. Stars: 634 total, peak 6 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 |
|---|---|---|---|---|
| bryce13950 | 1K | 881 | 117 | 18 |
| jlarson4 | 364 | 271 | 38 | 23 |
| degenfabian | 86 | 57 | 25 | 2 |
| Copilot | 18 | 0 | 0 | 11 |
| github-actions[bot] | 14 | 14 | 0 | 0 |
| speediedan | 13 | 0 | 2 | 6 |
| notoookay | 12 | 0 | 0 | 8 |
| brendanlong | 12 | 0 | 3 | 4 |
| mivanit | 11 | 0 | 5 | 4 |
| david-wei-01001 | 5 | 0 | 3 | 2 |
| chanind | 5 | 0 | 1 | 1 |
| shayansadeghieh | 4 | 0 | 4 | 0 |
| BorisTheBrave | 4 | 0 | 0 | 3 |
| taziksh | 3 | 0 | 0 | 3 |
| tuomaso | 3 | 0 | 0 | 1 |
| danra | 3 | 0 | 3 | 0 |
| CarlG0123 | 3 | 0 | 0 | 2 |
| evcyen | 3 | 0 | 3 | 0 |
| mntss | 3 | 0 | 2 | 0 |
| RecreationalMath | 3 | 0 | 2 | 1 |
Recent activity
Latest issues, pull requests and releases
- Pull request#1573Austin1serb2026-07-30 21:50
- Pull request#1548jlarson42026-07-28 03:45
- Issue comment#1497almogtavor2026-07-23 15:41Mistral: use fork-capable attention bridge (enable hook_attn_in / set_use_attn_in)
- Pull request#1476jlarson42026-07-01 14:58
- Issue comment#1453jlarson42026-06-30 00:42Use Qwen config context lengths
- Issue#1454jlarson42026-06-29 23:53Support T5Gemma2 (Gemma 3) Architecture
- Issue#1446dylanberens2026-06-27 22:46[Proposal] Add Audio Spectrogram Transformer (AST) Adapter (ASTForAudioClassification)
- Issue comment#1434jlarson42026-06-26 22:33feat: add NemotronH hybrid Mamba2-Transformer architecture adapter
- Pull request#1434mukund19852026-06-22 21:57
- Pull request#1426danra2026-06-21 00:49
- Pull request#1425danra2026-06-21 00:48
- Pull request#1423danra2026-06-20 05:49
- Issue comment#1377jlarson42026-06-15 19:46Add Gemma 4 architecture support to TransformerBridge
- Pull request#1378robbiebusinessacc2026-06-10 22:20
- Issue comment#1080RecreationalMath2026-06-10 05:29Import fails by default in colab
- Pull request#1373TensorCruncher2026-06-09 04:31
- Issue comment#1363TensorCruncher2026-06-06 00:20[Question] What is the point of a round-trip test for the Olmo adapter?
- Issue#1351jlarson42026-06-01 13:57[Proposal] Add SmolLM3 TransformerBridge architecture adapter
- Pull request#1335jlarson42026-05-27 20:56
- Pull request#1333RecreationalMath2026-05-27 10:53
- Issue comment#1302jlarson42026-05-26 20:28[Proposal] Additional Architecture Adapter tests
- Pull request#1329RecreationalMath2026-05-26 17:20
- Pull request#1309Rishik002026-05-21 04:27
- Issue comment#1309jlarson42026-05-20 21:35Adding adapter tests for Qwen2
- Issue comment#1263TravisHaa2026-05-20 01:35[Proposal] Direct Logit Attribution Tool
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 634 stars here means stars gained during the window, not the repo's star count.