Skip to content

Add sparsify error benchmarking artifacts#263

Open
noajshu wants to merge 1 commit into
quantumlib:mainfrom
noajshu:tesseract-benchmarking-code
Open

Add sparsify error benchmarking artifacts#263
noajshu wants to merge 1 commit into
quantumlib:mainfrom
noajshu:tesseract-benchmarking-code

Conversation

@noajshu

@noajshu noajshu commented Jun 18, 2026

Copy link
Copy Markdown
Contributor

Adds the benchmarking artifacts used for the error sparsification work in #254 under benchmarking/sparsify_errors/.

This includes:

  • the Slurm submission script used for the sparsify-errors sweep
  • the aggregated JSONL benchmark dataset
  • the plotting/analysis script used to generate the PR figures
  • the generated PDF plots

@noajshu noajshu marked this pull request as ready for review June 18, 2026 21:48
@noajshu noajshu requested a review from a team as a code owner June 18, 2026 21:48
@noajshu noajshu requested review from LalehB and arshpreetmaan and removed request for a team June 18, 2026 21:48

@arshpreetmaan arshpreetmaan left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

All good.

actual_path = alt_path
else:
_E_CACHE[circuit_path] = (1, 1)
return (1, 1)

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

when we are returning (1, 1) here when the circuit file is missing is it possible that it would generate plausible-looking but incorrect plots for some cases? wdyt of raising an error by default, or adding an explicit --allow-missing-circuits mode for offline reproduction?

_E_CACHE = {}

# Fallback mapping extracted directly from your provided output.
_FALLBACK_DETECT_MAP = {

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

For this fallback map that depends on exact circuit filenames. wdyt if we either compute these metrics from the circuit files directly or store E and num_detectors in aggregated_results.jsonl? That would make the analysis less coupled to one benchmark.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants