eval.qa · aif · dashboard

AIF dashboard

Append-only JSONL store. Refresh after submitting a form to see new records.

107
total evals
4.02
avg overall
0
violating safety
-
human vs LLM Δ
3
rater types

By rater type

human 100
end_user 6
llm_judge 1

By modality

text 1
chat 90
agent 1
rag 7
unknown 8

Recent evals

whenraterleveltemplatemodalitytaskoverallsafetytags
2026-07-11T00:09 [email protected] human enduser chat ad-hoc 2 clean
2026-07-11T00:08 [email protected] human enduser chat ad-hoc 1 clean
2026-07-09T16:14 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-09T13:36 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-09T12:46 [email protected] human enduser chat ad-hoc 4 clean helpful safe
2026-07-09T00:07 [email protected] human enduser chat ad-hoc 3 clean incomplete
2026-07-08T16:04 [email protected] human enduser chat ad-hoc 4 clean helpful
2026-07-08T03:54 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-08T01:15 [email protected] human enduser chat ad-hoc 2 clean
2026-07-07T15:00 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-07T14:58 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-07T14:31 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-07T13:16 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe +1
2026-07-07T12:39 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe +2
2026-07-07T12:36 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-07T12:15 [email protected] human enduser chat ad-hoc 3 clean off_topic incomplete
2026-07-07T10:06 [email protected] human enduser chat ad-hoc 3 clean
2026-07-07T07:45 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-07T07:24 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-07T07:23 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-07T07:19 [email protected] end_user - - - 5 clean
2026-07-07T07:15 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-07T06:59 [email protected] end_user - - - 4 clean
2026-07-07T06:24 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
2026-07-07T06:20 [email protected] human enduser chat ad-hoc 5 clean helpful accurate safe
+ New evaluation Eval Army profile Embed SDK stats JSON vision