Review economics
NVIDIA/cutlass
Read from the 100 most recently merged pull requests ·
Pull requests with attributed agent authorship took 1.0× the reviews submitted of the rest.
≥5%
Attributed
a floor
4
Attributed PRs
96
Other PRs
100
Read
Attributed — the latest 4
#3280 [SM120] Add ptr-array TMA collective for tensor/token-scaled FP8 grouped GEMM
1 round · 2 reviews · 17.0d to merge
#3070 [CuTeDSL] Add dataclass example: passing pointers via frozen dataclass
1 reviews · 32.6d to merge
#3047 docs: Fix float16 mislabeled as 32-bit in elementwise_add notebook
2 reviews · 18.4d to merge
#3004 [CuTeDSL] Add sub_packed_f32x2 operation
1 reviews · 8.9d to merge
The rest — the latest 5
#3394 Fix SM89 FP8 blockwise scale indexing under threadblock swizzling
1 reviews · 18.5d to merge
#3427 Fix ScaledBasis equality for mismatched bases, add SM120 sparse metadata K assert
2 reviews · 5.2d to merge
#3048 fix: Correct stride calculation in AffineRank2RowMajor::packed()
1 reviews · 165.4d to merge
#3244 [CuTe] Add missing include for smem_ptr_flag_bits in print_tensor.hpp
1 reviews · 80.0d to merge
#3432 v4.7 update.
1 reviews · 23h to merge
How this was measured
5% of merged commits carry agent attribution — a floor, not the share; tools that only complete code inline leave no commit trail, so the unattributed side includes AI-assisted work; reads 100 of 734 merged PRs — the rest have not been analyzed yet.
Detected: Claude Code, Cursor agent.
“Attributed” means a commit carried an agent’s signature — a co-author trailer, an agent commit identity, or an agent bot account. Tools that only complete code inline leave no such mark, so the other column is “rest”, not “human-written”. Full method and its limits
The badge reads this repo’s current report, so it follows the number.
Tell me when this moves
We re-read NVIDIA/cutlass weekly and email only when the number changes materially.
Read your own repos
This one is public. For a private repo, run the same read locally through your own GitHub credentials — nothing is installed and nothing is sent to us.
npx @ambera/review-taxWant this continuously on NVIDIA/cutlass — each pull request paired to the task it came from, and the work graded from its review loop rather than its diff size? Claim this repo in Forge
Browse every repo people have read · Read a different repository