experiment: run full-read Claude relevance baseline - #21

Closed

BestNathan wants to merge 19 commits into

mainfrom

experiment/full-read-baseline-20260924

Commits on Sep 24, 2026

experiment: compare single-file range and frontier v1b
BestNathan
committed
5cbe333
experiment: trigger single-file frontier v1b
BestNathan
committed
6d94f28
experiment: run balanced frontier v1c on agent websocket
BestNathan
committed
06911c2
fix: keep v1b experiment branch trigger
BestNathan
committed
27e0ee1
experiment: trigger balanced frontier v1c
BestNathan
committed
4ad2058
experiment: add full-read Claude relevance baseline workflow
BestNathan
committed
a3aade6
fix: pin frozen subject and stable baseline artifact name
BestNathan
committed
60d157b
experiment: trigger full-read Claude baseline
BestNathan
committed
f3baff1
fix: checkout baseline harness branch reliably
BestNathan
committed
fb623fe
experiment: retry full-read baseline
BestNathan
committed
e2383ae
fix: allow Claude Code proxy model for full-read baseline
BestNathan
committed
7097f20
experiment: retry full-read baseline with proxy model
BestNathan
committed
54674e9
fix: use Claude Code Sonnet for full-read reference baseline
BestNathan
committed
9ffb931
experiment: retry full-read baseline with Sonnet
BestNathan
committed
c50f542
fix: pin active Claude Sonnet 4.6 model for baseline
BestNathan
committed
a9046bc
experiment: retry full-read baseline with Sonnet 4.6
BestNathan
committed
98a5cb7
fix: run full-read CC baseline through ds model routing
BestNathan
committed
fd30d24
experiment: trigger full-read CC ds baseline
BestNathan
committed
c5769d9
fix: retry ds full-read baseline until canonical geometry validates
BestNathan
committed
0b9ca30

Navigation Menu

experiment: run full-read Claude relevance baseline - #21

experiment: run full-read Claude relevance baseline#21
BestNathan wants to merge 19 commits into
mainfrom
experiment/full-read-baseline-20260924

Commits on Sep 24, 2026

experiment: compare single-file range and frontier v1b

experiment: trigger single-file frontier v1b

experiment: run balanced frontier v1c on agent websocket

fix: keep v1b experiment branch trigger

experiment: trigger balanced frontier v1c

experiment: add full-read Claude relevance baseline workflow

fix: pin frozen subject and stable baseline artifact name

experiment: trigger full-read Claude baseline

fix: checkout baseline harness branch reliably

experiment: retry full-read baseline

fix: allow Claude Code proxy model for full-read baseline

experiment: retry full-read baseline with proxy model

fix: use Claude Code Sonnet for full-read reference baseline

experiment: retry full-read baseline with Sonnet

fix: pin active Claude Sonnet 4.6 model for baseline

experiment: retry full-read baseline with Sonnet 4.6

fix: run full-read CC baseline through ds model routing

experiment: trigger full-read CC ds baseline

fix: retry ds full-read baseline until canonical geometry validates