Help center
Interpret your Δ + TLX tiles in seconds
Use this cheat sheet before you paste Analyzer tiles into an exec deck or retro doc.
Calibrated (0–5)
Confidence ≈ evidence. Most self-ratings stay within ±5 of reviewer scores.
Keep logging reviewer minutes and sharing tiles in retrospectives.
Watch (5–15)
Storytelling is drifting from scored output. Execs will start asking for proof.
Pair high-Δ contributors with reviewers, inspect prompts, and check TLX in the same window.
Investigate (>15)
The narrative outruns evidence. Expect rework, compliance concerns, and blocked launches.
Freeze that workflow, review raw outputs, rerun the Fair Trial checklist, and reset the rubric.
TLX bands
Mental demand + frustration captured via the micro-TLX sliders.
Flow (0–40)
0–40
Team is relaxed and in control. Maintain cadence and share good practices.
Monitor (40–60)
40–60
Workload is rising. Discuss breaks, rotate reviewers, and ensure prompts are locked.
Elevated (60–80)
60–80
Fatigue is noticeable. Pause extra experiments and coach on System-2 checks.
Critical (80+)
80+
People are in fight-or-flight mode. Stop the run, review guardrails, and reset expectations.
Keep the story auditable
Link methodology for the math and quickstarts for context so stakeholders trust every tile. Review the methodology.
When Δ or TLX spike
Pause the run, share evidence, and reset guardrails before scaling.
- Re-run the Fair Trial checklist and lock prompts/rubrics.
- Compare reviewer minutes before and after the coaching block.
- Paste the updated tiles into the next exec memo so everyone sees the reset.
Δ & TLX FAQ
Answers to the three most common “what now?” questions.
What if Δ stays above 5 for a whole cohort?
Pause that workflow, review raw outputs with a reviewer, and reset prompts before logging more evidence.
How do TLX bands translate into actions?
40–60 triggers a coaching check-in, 60+ means pause the run and review guardrails, 80+ halts the workflow entirely.
Where should I share these tiles?
Paste them into every retro, exec memo, and changelog entry so stakeholders see Δ, TLX, and reviewer minutes together.
Open Beta
Help steer the Open Beta with real Δ and TLX tiles.
Run the analyzer demo, share methodology notes with your team, and send us benchmarks so the release ships with proof—not hype.
Next Steps
Ready to measure your AI impact? Start with a quick demo to see your Overestimation Δ and cognitive load metrics.