INDEX

Explanations

verbs or phrases indicating a planned or desired outcome

oai_token-act-pair · gpt-3.5-turbo

references to the purpose or intention behind actions, objects, or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

oner

-0.78

cit

-0.71

anon

-0.71

java

-0.69

visor

-0.68

tex

-0.68

SPONSORED

-0.68

acks

-0.65

mong

-0.64

 Videos

-0.63

POSITIVE LOGITS

 intended

0.78

 Parenthood

0.76

 beneficiaries

0.71

 recipients

0.68

IAL

0.67

 meant

0.63

ational

0.63

 provoking

0.62

 referen

0.62

ajor

0.61

Activations Density 0.022%

verbs or phrases indicating a planned or desired outcome

references to the purpose or intention behind actions, objects, or statements

No Comments

No Known Activations