INDEX

Explanations

instances of hypocrisy in actions versus stated beliefs

oai_token-act-pair · gpt-4o-mini Triggered by @bot

after vertical lines ("|") or formulas

np_acts-logits-general · gemini-2.0-flash

exam board abbreviations

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

vaadin

-0.58

STRUCTOR

-0.57

Aiheesta

-0.56

 saites

-0.56

structors

-0.53

Instructor

-0.53

 degradability

-0.53

 InputDecoration

-0.51

 instructors

-0.50

astéroïdes

-0.50

POSITIVE LOGITS

 governors

0.70

KS

0.65

 SEND

0.63

EY

0.61

SEN

0.58

SEND

0.58

sted

0.57

DIPSETTING

0.57

วย

0.55

Df

0.55

Activations Density 0.078%

instances of hypocrisy in actions versus stated beliefs

after vertical lines ("|") or formulas

exam board abbreviations

No Comments

No Known Activations