INDEX

Explanations

terms related to governmental bodies or organizations

oai_token-act-pair · gpt-3.5-turbo

references to governmental organizations or bodies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

sticks

-0.78

spring

-0.78

bound

-0.77

east

-0.75

Twe

-0.72

versions

-0.71

ammy

-0.69

Shop

-0.68

Phones

-0.68

QL

-0.67

POSITIVE LOGITS

governmental

1.07

vernment

0.98

ignty

0.88

gency

0.80

otonin

0.78

uthor

0.75

 organisations

0.75

assetsadobe

0.75

 governmental

0.74

 instit

0.73

Activations Density 0.005%

terms related to governmental bodies or organizations

references to governmental organizations or bodies

No Comments

No Known Activations