Agents
Per-agent hallucination risk and reliability — last 7 days. Each agent is tracked independently so regressions are immediately attributable.
hotpot-eval-runner
evaluation · 7d window · 0 traces
Avg reliability—
Avg hallucination risk—
Avg grounding—
Traces0
Hallucination risk over time
Daily average — lower is better. Spikes indicate degraded retrieval or synthesis.
No trend data in this window.
Reliability over time
Daily average reliability score — higher is better.
No reliability data yet.
squad-eval-runner
evaluation · 7d window · 0 traces
Avg reliability—
Avg hallucination risk—
Avg grounding—
Traces0
Hallucination risk over time
Daily average — lower is better. Spikes indicate degraded retrieval or synthesis.
No trend data in this window.
Reliability over time
Daily average reliability score — higher is better.
No reliability data yet.