fix: apply logit bias before greedy sampling#507
Merged
HenryNdubuaku merged 4 commits intomainfrom Mar 9, 2026
Merged
Conversation
Signed-off-by: Noah Cylich <[email protected]>
Contributor
There was a problem hiding this comment.
Pull request overview
Fixes sampling correctness by ensuring logit bias is applied before the greedy (temperature=0) early-return path, so greedy decoding selects the argmax of biased logits rather than raw logits.
Changes:
- Apply logit bias before the greedy early-return in
cactus_sample_f32andcactus_sample_f16. - Replace manual copy loops with
std::memcpyand manual argmax loops withstd::max_element. - In FP16 sampling, ensure temperature scaling operates on
filtered_logits(so bias is preserved through scaling).
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Moves the empty-vocab guard to the top of cactus_sample_f32 and cactus_sample_f16 so it covers all paths (not just greedy), preventing UB from std::max_element on empty vectors. Signed-off-by: Noah Cylich <[email protected]>
Signed-off-by: Noah Cylich <[email protected]>
Contributor
There was a problem hiding this comment.
Pull request overview
Copilot reviewed 2 out of 2 changed files in this pull request and generated 5 comments.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
assert(n <= 8) is compiled out in release builds, leaving the fixed-size float c[8] buffer unprotected. Replace with an early return. Signed-off-by: Noah Cylich <[email protected]>
ccb37f5 to
4fab757
Compare
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
cactus_sample_f32andcactus_sample_f16std::max_elementand manual copy loops withstd::memcpy