INDEX

Explanations

names of specific individuals or characters

oai_token-act-pair · gpt-3.5-turbo

proper nouns related to individuals and places

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ãĥīãĥ©

-0.73

Vi

-0.69

emic

-0.69

cs

-0.67

 Ãĸ

-0.67

NAT

-0.66

 crystal

-0.65

osterone

-0.64

Az

-0.64

Uz

-0.64

POSITIVE LOGITS

 Gregg

3.21

 Schneider

1.64

Gre

1.63

 Kerr

1.55

 Greene

1.23

 Greenberg

1.21

 Jenkins

1.18

 Ferry

1.15

 Greenpeace

1.15

 Prest

1.13

Activations Density 0.044%

names of specific individuals or characters

proper nouns related to individuals and places

No Comments

No Known Activations