Skip to content

Pull requests: karpathy/autoresearch

New pull request New
85 Open 102 Closed
85 Open 102 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

docs: guide for evaluating experiment results at scale
#303 opened Mar 16, 2026 by dean0x Loading…
Karpathy's AutoResearch with Memory-in-the-Loop States by The Adimension
#302 opened Mar 16, 2026 by habanwer Loading…
8
Working version with nvidia geforce 1050 ti 4GB
#299 opened Mar 16, 2026 by ssingh10 Loading…
Fix: add data integrity verification to downloader to prevent corrupted shardsfeat: add data integrity verification to downloader
#291 opened Mar 16, 2026 by GhostDragon124 Loading…
Fix typos and improve readability in README.md
#290 opened Mar 16, 2026 by jmfinch Loading…
1
Train: add gradient clipping before optimizer step
#287 opened Mar 16, 2026 by renee-jia Loading…
Bake reflection into the experiment loop
#282 opened Mar 15, 2026 by alexbenari Loading…
3
feat: add --profile flag for LLM-readable CUDA kernel summary
#279 opened Mar 15, 2026 by Bortlesboat Loading…
feat(autoresearch): Implement keep/discard policy engine
#276 opened Mar 15, 2026 by nmandal Loading…
fix: respect HF_ENDPOINT env var for dataset downloads
#272 opened Mar 15, 2026 by ycjcl868 Loading…
2 tasks
feat: Turing GPU (T4/RTX) compat via capability-aware FA3 fallback, fp16+GradScaler
#268 opened Mar 14, 2026 by Nidhicodes Loading…
2 tasks done
save checkpoint before eval to survive crashes
#265 opened Mar 14, 2026 by mvanhorn Loading…
1
4
fix(train): guard against zero grad_accum_steps
#264 opened Mar 14, 2026 by jluk Loading…
fix: define 'best' variable before use in progress plot
#263 opened Mar 14, 2026 by mango766 Loading…
1
docs: minimal CONTRIBUTING.md focused on research context
#258 opened Mar 14, 2026 by guoyangzhen Loading…
Add autoresearch-webgpu to notable forks
#253 opened Mar 14, 2026 by lucasgelfond Loading…
Add autoresearch-mac-mini to Notable Forks
#252 opened Mar 14, 2026 by Venkateshwar-PortoAI Loading…
feat(benchmarks): MPS vs MLX Benchmark Suite + stdio_bus Documentation
#248 opened Mar 13, 2026 by morozow Loading…
autotrader/mar13: GBR model with val_pass=true (score 1.40)
#245 opened Mar 13, 2026 by ijpatter1 Loading…
2 tasks
tune: apply discussion-43 final config to master
#244 opened Mar 13, 2026 by dan-y Loading…
docs: add Codex looping note
#241 opened Mar 13, 2026 by Bortlesboat Loading…
Harden tokenizer deserialization to prevent cache poisoning RCE
#237 opened Mar 13, 2026 by Mirza-Samad-Ahmed-Baig Loading…
feat(stdio_bus): unified CUDA/MPS GPU backend support
#230 opened Mar 13, 2026 by morozow Loading…
Add MLX port for MacOS (optimized for 16GB RAM) to notable forks
#227 opened Mar 12, 2026 by matt-k-wong Loading…
feat(stdio_bus): Add stdio_bus Integration for Distributed GPU Training
#223 opened Mar 12, 2026 by morozow Loading…
ProTip! Updated in the last three days: updated:>2026-03-14.