INDEX

Explanations

references to the word "book."

oai_token-act-pair · gpt-3.5-turbo

the word "ok" 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

lav

-0.80

 Legions

-0.74

 cort

-0.72

FACE

-0.67

Engineers

-0.67

 draped

-0.67

 Veil

-0.65

âĢ¢âĢ¢âĢ¢âĢ¢

-0.64

 duct

-0.63

ãĥĺãĥ©

-0.62

POSITIVE LOGITS

lahoma

1.17

ok

1.09

arak

1.01

wana

0.95

uten

0.93

oro

0.92

sha

0.90

owski

0.89

obo

0.89

awaru

0.89

Activations Density 0.012%

references to the word "book."

the word "ok" in various contexts

No Comments

No Known Activations