INDEX

Explanations

references to the term "WW" or variations like "WWF" and "WWII"

oai_token-act-pair · gpt-3.5-turbo

references to World War events and organizations

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

Downloadha

-0.96

alties

-0.85

Reviewer

-0.83

uated

-0.81

staking

-0.79

 succeeding

-0.74

eca

-0.71

holders

-0.71

 fitting

-0.69

uate

-0.69

POSITIVE LOGITS

WW

1.30

JD

0.85

WF

0.83

restling

0.82

AMP

0.81

ombat

0.79

II

0.79

ITNESS

0.78

 Norton

0.76

EST

0.76

Activations Density 0.007%

references to the term "WW" or variations like "WWF" and "WWII"

references to World War events and organizations

No Comments

No Known Activations