Try Hanzo

Benchmarks / Kai / Support triage, validation split

Kai · Validation split

Support triage, validation split

Kai on Support triage, validation split, one run per revision. No baseline was run on this split.

Latest

History

6 results across 2 metrics, newest first. Superseded runs stay, so a revision's progress can be read.

64.6%65.6%66.6%09-29Kai triage+job_30bfc6e5af09a9f7 · 0cac9e85: 64.9%, Sep 29, 2026Kai a7 · 0834a74f: 66.4%, Sep 29, 2026Kai triage+job_cccd2fcfeae9ecb9 · 57535ef4: 65.8%, Sep 29, 2026
MeasuredSubjectMetricValueBaselineResultStatus
Sep 29, 2026Kai triage+job_30bfc6e5af09a9f7 · 0cac9e85Accuracy64.9%–UnscoredLive · Hanzo-measured
Sep 29, 2026Kai a7 · 0834a74fAccuracy66.4%–UnscoredSuperseded · Hanzo-measured
Sep 29, 2026Kai triage+job_cccd2fcfeae9ecb9 · 57535ef4Accuracy65.8%–UnscoredSuperseded · Hanzo-measured
Sep 29, 2026Kai triage+job_30bfc6e5af09a9f7 · 0cac9e85Calibration error (ECE)0.282–UnscoredLive · Hanzo-measured
Sep 29, 2026Kai a7 · 0834a74fCalibration error (ECE)0.262–UnscoredSuperseded · Hanzo-measured
Sep 29, 2026Kai triage+job_cccd2fcfeae9ecb9 · 57535ef4Calibration error (ECE)0.278–UnscoredSuperseded · Hanzo-measured

Conditions

split: val:0b8b6476822b343c518 questions, 518 answereddataset sha256:0b8b63 Kai revisions measured

Hardware

ROCm BF16 on evo

Source

research run kai.support_triage-val:0b8b6476822b343c-triage+job_30bfc6e5af09a9f7.0cac9e85-kai

Reproduce

No public command regenerates this result. The source above is the record it is read from.

More Kai benchmarks

AG News · DAIR Emotion · Banking77 · Support triage · Email spam · Phishing · Jailbreak · Toxicity · RAG relevance · Model routing · Typed decisions · MASSIVE · AG News, validation split · Typed decisions, validation split · HarmBench (benign prompts) · Choosing among many options · Joint decisions · Release gate against Jev · Latency of one decision

Every result for Support triage, validation split on the shelf · How these are measured

Build what’s next.