Promote hand-written 2017 notes out of Archive/ into notes/#7
Open
0bserver07 wants to merge 1 commit into
Open
Promote hand-written 2017 notes out of Archive/ into notes/#70bserver07 wants to merge 1 commit into
0bserver07 wants to merge 1 commit into
Conversation
The trusted, hand-written CS294 notes were sitting in a folder called Archive — which sounds like "old/dead" — while the unreviewed AI-drafted lecture series occupied notes/. Backwards from a learner's perspective. - Archive/2017-Course-Notes/CS294-DeepRL-Berkeley/ -> notes/cs294-2017/ (with imgs/ intact) - Archive/2017-Course-Notes/Elements-Of-RL/ -> notes/sutton-barto-digest/ - Both files got <!-- status: hand-written --> headers - Archive/ directory deleted (Archive/README.md was just a wrapper) readme.md restructured: "What's here" now leads with trusted hand-written content (the CS294 notes, the Sutton & Barto digest, the curated talks/books/courses, the tested exercises). The AI-drafted lecture series is clearly demoted as "scaffold, treat with skepticism." "Start here" reordered: talks/books -> exercises -> drafts. notes/README.md rewritten in the same spirit. AGENTS.md and CLAUDE.md updated to point at notes/cs294-2017/ and notes/sutton-barto-digest/ as the trusted, frozen, never-reword material. GitHub topics refreshed separately (not in this commit): dropped `guideline` and `study`; added `rlhf`, `llm-alignment`, `dpo`, `grpo`, `ppo`, `rlvr`, `agentic-rl`, `lecture-notes`, `study-notes`, `deepseek-r1`, `constitutional-ai`, `policy-gradient`, `q-learning`, `sutton-barto`. Description sharpened. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
You were right — the structure had the trusted, hand-written CS294 notes filed under
Archive/(which sounds like a museum) while the unreviewed AI-drafted lectures occupiednotes/(the "main" content). For someone learning RL from this repo, that's exactly backwards: the part they can actually trust is buried, the part they should treat with skepticism is featured.This PR fixes it. Not merging — you review and tell me to merge or change something.
What moved
Archive/2017-Course-Notes/CS294-DeepRL-Berkeley/→notes/cs294-2017/(withimgs/intact, image links unchanged)Archive/2017-Course-Notes/Elements-Of-RL/→notes/sutton-barto-digest/Archive/README.mddeleted (it was a wrapper; no content lost)Archive/directory goneBoth moved files got a
<!-- status: hand-written -->header so they're explicitly marked as trusted.What changed in the docs
readme.md"What's here" now leads with Trusted, hand-written (CS294 notes, Sutton & Barto digest, curated talks/books/courses, tested exercises), then AI-drafted, useful as scaffold (treat with skepticism) (the 19-lecture series, cheat sheets, diagrams, scraped paper lists, tools).readme.md"Start here" reordered: talks/books/courses first → exercises → drafts → AGENTS.md.notes/README.mdrewritten the same way — leads with the hand-written content, the lecture-series table sits below clearly framed as drafts.AGENTS.mdandCLAUDE.mdupdated: the layout table now points atnotes/cs294-2017/andnotes/sutton-barto-digest/as the trusted, frozen, never-reword material instead ofArchive/. The voice rule that referenced "the 2017 archive" updated to point at the new path.What I did separately (already live, not in this PR)
GitHub topics refreshed: dropped
guidelineandstudy; addedrlhf,llm-alignment,dpo,grpo,ppo,rlvr,agentic-rl,lecture-notes,study-notes,deepseek-r1,constitutional-ai,policy-gradient,q-learning,sutton-barto. Description sharpened to "RL study guide — foundations through RLHF, DPO, GRPO, RLVR, agentic RL, and offline RL. Hand-written CS294 notes, 19 lecture drafts, 5 tested exercises, citations that resolve."🤖 Generated with Claude Code