INDEX

Explanations

proper nouns or names of individuals

oai_token-act-pair · gpt-3.5-turbo

names of people and entities significant in political contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Carter

-0.75

onday

-0.74

aid

-0.70

 enterprise

-0.68

lease

-0.67

FACE

-0.66

 Lamar

-0.66

 loft

-0.65

 interchange

-0.65

eway

-0.64

POSITIVE LOGITS

 Bers

3.65

 Grimm

1.61

Hau

1.57

 Grim

1.49

Anim

1.42

Gau

1.37

Urs

1.29

 Berserker

1.14

Gry

0.99

 Anim

0.97

Activations Density 0.054%

proper nouns or names of individuals

names of people and entities significant in political contexts

No Comments

No Known Activations