INDEX

Explanations

phrases related to sports fields or battlefields

oai_token-act-pair · gpt-3.5-turbo

references to territorial conflicts or battle grounds

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

legraph

-0.79

utra

-0.75

uras

-0.74

essee

-0.70

alez

-0.69

ems

-0.68

rophe

-0.68

ustainable

-0.67

izations

-0.67

awed

-0.66

POSITIVE LOGITS

 turf

1.23

grass

0.90

Wiz

0.74

mint

0.72

toe

0.72

lore

0.69

Zar

0.69

ilage

0.68

lette

0.68

FIELD

0.68

Activations Density 0.006%

phrases related to sports fields or battlefields

references to territorial conflicts or battle grounds

No Comments

No Known Activations