Anti-modal yet feasibleT-0.42
Where creative research escapes mode collapse

DIVERGA

Beyond predictable AI recommendations. Diverge from the obvious.

24 research agents preventing mode collapse through VS methodology.

24 Agents·Specialized Research
VS Methodology·Beyond Mode Collapse
T-Score·Typicality-Aware
Claude Code·Exclusive Platform
Scroll
T-Score System

The T-Score Spectrum

Most AI outputs cluster at T-Score 0.8+ (modal). Diverga helps you explore the creative long tail.

Divergent (0.0)(0.5)Modal (1.0)
0.92

TAM

T-0.92

0.85

UTAUT

T-0.85

0.55

Hybrid

T-0.55

0.38

SDT+TAM

T-0.38

0.18

Novel

T-0.18

VS Methodology

VS Methodology in Action

One query, five directions. Human decides which path to explore.

Research Query
0.150.350.500.720.92
Divergent
Creative
Balanced
Typical
Modal
DivergentT-0.15

Challenge existing assumptions with counter-intuitive approaches

Examples
Paradoxical relationshipsAssumption-challenging hypotheses
Recommendations

Best for paradigm-shifting research with strong theoretical justification

Click to explore
24 Agents

Agent Collaboration Network

24 specialized agents work together across 9 research lifecycle categories.

A: Foundation
B: Evidence
C: Design & Meta
D: Data Collection
E: Analysis
F: Quality
G: Communication
H: Specialized
24 Agent Network
Try It Now

Interactive Terminal

Experience Diverga CLI with human checkpoints.

diverga-cli — terminal
DIVERGA CLI v9.0 - Research Methodology AI Assistant Connected: Groq LLM (llama-3.3-70b) Ask anything! Examples: "What is T-Score?" "Help me refine my research question" "How do I design a meta-analysis?" Commands: help | agents | run <agent_id>
$

Ready to Escape Mode Collapse?

Join researchers who are exploring the long tail of methodology.

This documentation: T-Score 0.42 (Creative Range)