INDEX

Explanations

words related to political figures or entities

oai_token-act-pair · gpt-3.5-turbo

proper nouns associated with notable entities or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

stration

-0.71

legram

-0.68

aunder

-0.67

amina

-0.67

hess

-0.67

ancial

-0.67

ript

-0.67

enda

-0.66

vironment

-0.66

chet

-0.66

POSITIVE LOGITS

 deems

0.79

 deem

0.70

 Enterprises

0.68

 chose

0.67

 chooses

0.66

 stole

0.66

 describes

0.64

 assigns

0.64

 fans

0.63

 refers

0.63

Activations Density 0.186%

words related to political figures or entities

proper nouns associated with notable entities or groups

No Comments

No Known Activations