loupe / reports

46.8 kB

Ctrl+K

1 contributor

History: 5 commits

anxuanng

Gate chart: one-line note

5fd9805 verified 4 days ago

benchmark.json

10.1 kB
Loupe research preview: audit code, read weights, benchmark reports 5 days ago
design2code_human_pairwise.json

1.42 kB
Add the Design2Code human-judged benchmark: Loupe does not beat a pixel baseline there 4 days ago
design2code_similarity.json

1.69 kB
Continuous similarity: 76.4 percent agreement with human judges on held-out Design2Code pairs 4 days ago
design2code_sonnet.json

197 Bytes
Gate chart: one-line note 4 days ago
first_24_sites.json

7.53 kB
Loupe research preview: audit code, read weights, benchmark reports 5 days ago
live_refit_rejected.json

1.46 kB
Loupe research preview: audit code, read weights, benchmark reports 5 days ago
multi_line_reading.json

458 Bytes
Loupe research preview: audit code, read weights, benchmark reports 5 days ago
repair_suite.json

2.07 kB
Gate chart: one-line note 4 days ago
sensitivity.json

1.01 kB
More benchmarks: eight windows per fault against Sonnet, fault size test, speed, charts 4 days ago
timing.json

1.9 kB
More benchmarks: eight windows per fault against Sonnet, fault size test, speed, charts 4 days ago
unseen_32_pages.json

5.24 kB
Loupe research preview: audit code, read weights, benchmark reports 5 days ago
unseen_33_pages.json

5.65 kB
Loupe research preview: audit code, read weights, benchmark reports 5 days ago
versus_sonnet.json

3.04 kB
More benchmarks: eight windows per fault against Sonnet, fault size test, speed, charts 4 days ago
versus_sonnet_first_4_per_fault.json

3.04 kB
More benchmarks: eight windows per fault against Sonnet, fault size test, speed, charts 4 days ago
versus_sonnet_prompt.txt

2.01 kB
Loupe research preview: audit code, read weights, benchmark reports 5 days ago