INDEX

Explanations

mentions of a specific term "Boy" within the text

oai_token-act-pair · gpt-3.5-turbo

references to the term "Boy" in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aeda

-1.01

aukee

-0.75

 podium

-0.73

Ļ½

-0.72

anwhile

-0.72

 equilibrium

-0.70

 totality

-0.70

 instability

-0.69

incinn

-0.69

srf

-0.69

POSITIVE LOGITS

friend

1.39

boy

1.28

Boy

1.24

Boy

1.24

 Scouts

1.20

cott

1.12

boys

1.10

 Scout

1.02

hood

1.00

Girl

0.99

Activations Density 0.006%

mentions of a specific term "Boy" within the text

references to the term "Boy" in various contexts

No Comments

No Known Activations