Persona Evaluation — condition5__discord_only__sample500_seed99

Field	Value
Persona	Nimesh Kulatunga
Judge model	openai/gpt-4.1-mini
Embed model	text-embedding-3-small
Rubric draws (k)	3
Total suggestions	50
Pipeline mode	hall_pass=skipped (active=False)

Bucket Distribution

Bucket	Count	% of total
Task Critical	16	32.0%
Quality Of Life	24	48.0%
Noise	10	20.0%

Set-Level Diversity Metrics

Metric	Value	Interpretation
DPP log-det	-40.6520	Higher = more diverse + high-quality set
Cluster coverage	1.000	Fraction of BGT clusters with a task-critical hit
ILAD	0.6100	Mean pairwise distance; higher = more diverse
Redundancy rate	0.000	Fraction of near-duplicate suggestions (cos > 0.9)

Composite Score

Component	Weight	Value
DPP set score (normalised)	0.5	—
Cluster coverage	0.3	1.000
Mean quality (non-hallucinated)	0.2	—
Hallucination penalty	alpha=0.5	x 1.0000

Composite score: 0.6115

Hallucination Summary

Filter: skipped. Hallucination pass skipped (no anti-GT). Penalty pinned to 1.0 so the composite formula and weights stay identical to the active mode and scores remain comparable across runs.

Filtered Hallucinations

Filter inactive — no suggestions were inspected for anti-GT hallucinations.

Top 5 Task-Critical Suggestions

#	ID	Quality	Title
1	22	1.000	Apply Proposal Formatting for April 25 Deadline
2	9	0.970	Prepare for the 'Proof of Work' Application Phase
3	14	0.970	Apply 'Strategic Under-Crowding' to Project Selection
4	21	0.970	Pre-Verify Stipend Wallet Requirements
5	25	0.970	Deep-Dive into Coinswap and Mercury Layer

Top 5 Quality-of-Life Suggestions

#	ID	Quality	Title
1	16	0.970	Prepare Round 2 Pivot List by April 11
2	11	0.947	Secure Portal Access and Profile Optimization Steps
3	18	0.937	Apply 'High-Demand' Strategy for Round 1 Selection
4	48	0.937	Prepare for Round 2 Registration (April 12–14)
5	30	0.927	Project Selection Strategy: The '2-1-1' Method