Spaces:

Nomearod
/

agentbench

Running

App Files Files Community

agentbench / measurements

123 kB

Ctrl+K

Ctrl+K

4 contributors

History: 5 commits

Nomearod's picture

calibrate(jury): 4A characterizes v1.1.1 residual as model-class-specific

504a35c 25 days ago

2026-04-15-coldstart-n1.log

2.16 kB
docs: cold-start measurement + falsified-assumption finding + v1.1 contingency about 2 months ago
2026-04-15-coldstart-n2.log

2.14 kB
docs: cold-start measurement + falsified-assumption finding + v1.1 contingency about 2 months ago
2026-04-15-coldstart-n3.log

2.17 kB
docs: cold-start measurement + falsified-assumption finding + v1.1 contingency about 2 months ago
2026-05-04-judge-calibration-labels.jsonl

26.4 kB
rubric: clarify groundedness reference scope (snippets-only) for v1.1 gold (#20) 26 days ago
2026-05-05-judge-rubric-opus-stress.jsonl

72.4 kB
rubric: clarify groundedness reference scope (snippets-only) for v1.1 gold (#20) 26 days ago
2026-05-06-3a-paraphrase-recency-probe.jsonl

4.11 kB
calibrate(jury): v1.1+v1.1.1 — fix weighting bugs; recency-position paraphrase clause 25 days ago
2026-05-06-4a-gpt4o-full-probe.jsonl

5.1 kB
calibrate(jury): 4A characterizes v1.1.1 residual as model-class-specific 25 days ago
2026-05-06-gpt4o-extraction-reasoning-split.md

7.66 kB
calibrate(jury): v1.1+v1.1.1 — fix weighting bugs; recency-position paraphrase clause 25 days ago
README.md

1.05 kB
docs+build: judge-layer v1 coupled-artifact updates 27 days ago