gitzette / @karpathy / Mar 2 – Mar 8 W10
← W09 W10 W11 →
@karpathy — open-source digest March 2 – March 8, 2026
Vol. 1, No. 0 github.com/karpathy
the dispatch
@karpathy
Quiet week on GitHub. 170 commits across the autoresearch universe. No releases.
170 commits 0 PRs 0 releases 10 repos

170 commits spread across 10 repos, mostly autoresearch variants. No releases, no merged PRs, no fires to report.

PENDING

autoresearch stuck in the issues queue

Five open issues outline the next frontier: licensing questions, pluggable optimization tracks, and a pivot toward data-centric research.

@karpathy has filed five substantial issues against autoresearch this week, none yet merged. The most ambitious: Auto-Research Tracks, which proposes pluggable idle optimization across harness, skills, memory, and model selection. A separate pitch for data-centric autoresearch suggests the project may be at an inflection point. Pending work also includes a JudgeModel layer for snapshot-diff scoring and a call to expand the harness generator into the pattern/memory/security design space. Licensing remains an open question.

COMMUNITY

autoresearch-cpu accumulates commits

The CPU-bound variant of autoresearch picked up 17 commits this week but remains undocumented in the public record.

@karpathy's CPU adaptation saw steady churn. No releases, no PRs, no issues filed. The work is opaque from the outside.

COMMUNITY

EvolvedAI keeps the lights on

The social simulation fork of autoresearch registered 17 commits but no user-facing activity.

@karpathy committed to the prompt-only social simulation built from autoresearch, but nothing shipped or broke this week.

the dispatch — Vol. 1, No. 0 March 2 – March 8, 2026
COMMUNITY

autokernel5090 waits for the RTX 5090

GPU-specific kernel work accumulated 17 commits with no visible release or PR activity.

@karpathy is preparing infrastructure for the RTX 5090. The commits landed but stayed under wraps. Likely not production-ready yet.

COMMUNITY

TrainRTX5070 training pipeline moves quietly

Continual learning setup for RTX 5070 saw 17 commits this week with no releases or issues filed.

@karpathy's RTX 5070 training pipeline accumulated work in silence. No public signals yet on what the commits contain or when they ship.

stats corner
170
commits
0
pull requests
0
releases
commits by repo
REPO COMMITS autoresearch 17 autoresearch 17 EvolvedAI 17 autoresearch 17 apex_mes_reasearch 17 Jessi 17 autokernel5090 17 autoresearch-cpu 17 autoresarch-macos-poc 17 TrainRTX5070 17
github stars
autoresearch
48,246
TrainRTX5070
1
repo index
  • autoresearch
    AI agents running research on single-GPU nanochat training automatically
    17 commits · 0 releases · ★ 48246
  • autoresearch
    Autonomous pretraining research swarm — adapted for Mac (MPS) and T4 GPU
    17 commits · 0 releases · ★ 0
  • EvolvedAI
    AutoCiv / EvolvedAI prompt-only social simulation built from karpathy/autoresearch
    17 commits · 0 releases · ★ 0
  • autoresearch
    17 commits · 0 releases · ★ 0
  • apex_mes_reasearch
    17 commits · 0 releases · ★ 0
  • Jessi
    17 commits · 0 releases · ★ 0
  • autokernel5090
    AutoKernel for 5090 GPU
    17 commits · 0 releases · ★ 0
  • autoresearch-cpu
    CPU adaptation of Karpathy's autoresearch — autonomous ML research on commodity hardware, no GPU required
    17 commits · 0 releases · ★ 0
  • autoresarch-macos-poc
    autoresarch-macos-poc
    17 commits · 0 releases · ★ 0
  • TrainRTX5070
    Continual learning pipeline for RTX5070 training the maximum optimal model with it.
    17 commits · 0 releases · ★ 1
Mar 9 – Mar 15 →
gitzette @karpathy on gitzette @karpathy on GitHub
share: post on X share on LinkedIn

Your GitHub week, turned into something worth reading.

Generate your dispatch →
gitzette.online  ·  2026 © AISlopMedia, Inc.