INDEX

Explanations

action or behavior words ending in '-ishing' and '-ished'

oai_token-act-pair · gpt-3.5-turbo

words related to different forms of punishment or negative consequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

udi

-0.69

 reorgan

-0.69

agonists

-0.67

AMY

-0.67

sen

-0.65

 autonom

-0.64

 Merrill

-0.64

ilogy

-0.63

oples

-0.63

imil

-0.63

POSITIVE LOGITS

ishes

0.95

ished

0.83

ments

0.83

ishing

0.81

ishment

0.81

âĶĢâĶĢ

0.79

ishers

0.78

manship

0.78

 Royale

0.78

ishly

0.73

Activations Density 0.029%

action or behavior words ending in '-ishing' and '-ished'

words related to different forms of punishment or negative consequences

No Comments

No Known Activations