INDEX

Explanations

instances of the word 'and'

oai_token-act-pair · gpt-3.5-turbo

the word "and" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

zar

-0.77

Untitled

-0.72

zag

-0.72

shift

-0.69

Ñı

-0.68

icter

-0.67

ident

-0.67

du

-0.67

NULL

-0.66

POSITIVE LOGITS

 although

1.21

 moreover

1.21

 consequently

1.20

 furthermore

1.20

 therefore

1.08

ersen

1.03

 hence

1.02

 rightfully

1.00

 thus

0.99

 rightly

0.94

Activations Density 0.275%

instances of the word 'and'

the word "and" in various contexts

No Comments

No Known Activations