INDEX

Explanations

the word "willing" in the text

oai_token-act-pair · gpt-3.5-turbo

expressions that indicate a willingness or readiness to take action

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

gran

-0.78

 Anthem

-0.72

Pt

-0.69

lav

-0.69

ORGE

-0.68

alien

-0.68

cell

-0.67

adish

-0.66

ogg

-0.66

gar

-0.66

POSITIVE LOGITS

 willing

1.22

 unwilling

1.05

theless

0.88

 incent

0.84

 willingly

0.82

 onlook

0.78

 unres

0.76

 gladly

0.76

 compe

0.75

bilt

0.74

Activations Density 0.014%

the word "willing" in the text

expressions that indicate a willingness or readiness to take action

No Comments

No Known Activations