INDEX

Explanations

phrases related to authority or importance

oai_token-act-pair · gpt-3.5-turbo

frequent use of the article "the."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

âĢł

-0.80

cum

-0.76

ãĤ´ãĥ³

-0.75

udder

-0.69

:-

-0.68

 nonetheless

-0.68

.(

-0.68

ãĥ»

-0.67

 thereby

-0.67

.-

-0.67

POSITIVE LOGITS

 whole

1.08

 totality

1.01

 biggest

1.00

 hardest

0.98

 toughest

0.97

guy

0.96

oret

0.93

 greatest

0.93

 slightest

0.92

Activations Density 0.492%

phrases related to authority or importance

frequent use of the article "the."

No Comments

No Known Activations