Add kernelbench-style annotation for opus_18.45x record - #1
Open
AlexWortega wants to merge 1 commit into
Open
Conversation
Documents the OpenRSI (Opus 4.8) 18.45x Kimi-Linear W4A16 decode megakernel in kernelbench.com's annotation schema, with links to the kernel and the RSI scaffold. Off-GPU re-verification against kernelbench check.py logic: forbidden- import AST scan PASS and load_state_dict(strict=True) PASS (147/147 params match reference.py). Correctness (cosine >= 0.98) and 18.45x timing are OpenRSI's RTX PRO 6000 numbers, not re-run here. Co-Authored-By: Claude Opus 4.8 (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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds
mega_results/annotations/EXTERNAL_openrsi_claude-opus-4-8_02_kimi_linear_decode.yaml— a kernelbench.com-schema annotation for the OpenRSI (Opus 4.8) 18.45× Kimi-Linear W4A16 decode megakernel.English blurb (as requested): "My Recursive Self-Improvement agent on Opus 4.8 drove itself to this kernel for the RTX PRO 6000 (a6000pro)." Links to the kernel and the RSI scaffold.
Verification done (off-GPU, against kernelbench
check.pylogic):torch+load_inline, noimport reference, no banned libs.load_state_dict(strict=True)— all 147 params/buffers matchreference.pyby name and shape.Model.step(hidden, state) -> (hout, state)interface + singlemega_launchper step.Not re-run here (needs RTX PRO 6000): the cosine >= 0.98 correctness and the 18.45x timing remain OpenRSI's reported numbers. Annotation is labelled
verdict: interestingand flagged as an external, non-native-harness claim per kernelbench'sSCHEMA.md.🤖 Generated with Claude Code