INDEX

Explanations

proper nouns or names that end with "-amer"

oai_token-act-pair · gpt-3.5-turbo

references to American identity or context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ŃĶ

-0.80

ĵĺ

-0.73

erest

-0.68

fax

-0.65

margin

-0.64

pressed

-0.63

ird

-0.62

WAR

-0.61

 fermentation

-0.61

ģ«

-0.61

POSITIVE LOGITS

icans

1.26

ican

1.24

ICAN

1.19

gdala

1.03

icity

0.94

icas

0.93

ica

0.92

gency

0.90

ilial

0.89

agic

0.80

Activations Density 0.012%

proper nouns or names that end with "-amer"

references to American identity or context

No Comments

No Known Activations