INDEX

Explanations

words related to elementary school

oai_token-act-pair · gpt-3.5-turbo

references to elementary schools

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

cific

-0.81

 Archdemon

-0.78

-0.71

 palms

-0.70

————————————————

-0.67

atche

-0.65

reens

-0.65

Quantity

-0.65

igslist

-0.64

=-=-=-=-=-=-=-=-

-0.64

POSITIVE LOGITS

 school

0.97

 elementary

0.94

chool

0.88

school

0.85

 Elementary

0.84

 School

0.81

 Teachers

0.80

 schools

0.78

 SCHOOL

0.78

School

0.75

Activations Density 0.022%

words related to elementary school

references to elementary schools

No Comments

No Known Activations