MORGIN.AI researching the edge of AI behavior Research Benchmarks

Researching the edge of AI behavior.

We map how LLM systems fail under pressure - before those failures become expensive. For teams who cannot afford blind spots.

Local Models

How local models perform across hardware tiers, context demands, and sustained workloads.

Guardrails

What internal and layered safety systems catch, miss, and shift in production behavior.

Private Inference

How private-mode and anonymized routes differ, where data still flows, and what teams should verify before rollout.

Benchmark Library

Methods · detailed benchmark specs

EpsteinBench

EpsteinBench measures whether a model can continue a manipulative social thread in a way that is mistaken for the real archived reply.

PersuasionForGood Transfer Check

This benchmark reuses the EpsteinBench evaluation logic on human fundraising dialogue to test whether the adapter transfers something broader than archive-specific style.

Responsibility Avoidance

Responsibility Avoidance is a synthetic honesty stress test that asks what a model does when truthful disclosure becomes socially expensive.

+1 more benchmark See the full benchmark library

Recent Publications

Preview · latest three briefs

March 2026 · Benchmark Design

EpsteinBench: We Brought Epstein's Voice Back. We Got More Than We Wanted.

We trained a LoRA to capture Epstein's voice. The more disturbing change was in how the model pursued influence.

Ablation versus Heretic article banner image provided by user

March 2026 · Uncensoring Methods

Ablation vs Heretic vs Obliteratus: one trick, three layers of tooling

Ablation, Heretic, and Obliteratus are closely related. The real differences come down to how much tuning, tooling, and workflow each one adds.

Methods, outcomes, and mitigations documented clearly and updated regularly. Whitehat research for teams learning how local models and guardrails behave in practice.