INDEX

Explanations

names that start with "Mil" followed by a high activation value

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the name "Mil."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 compr

-0.89

OPLE

-0.84

chnology

-0.76

BOOK

-0.75

LY

-0.72

eering

-0.69

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.67

ĵĺ

-0.67

IELD

-0.66

ħĭ

-0.65

POSITIVE LOGITS

waukee

1.10

estones

1.06

ksh

1.06

estone

1.03

isec

1.03

gram

0.97

itary

0.93

ieu

0.92

ilit

0.92

itant

0.91

Activations Density 0.011%

names that start with "Mil" followed by a high activation value

repeated mentions of the name "Mil."

No Comments

No Known Activations