INDEX

Explanations

locations or points of reference within a text

oai_token-act-pair · gpt-3.5-turbo

phrases indicating locations or contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 hereby

-0.67

 favour

-0.63

morph

-0.61

robe

-0.60

 swear

-0.58

 Trade

-0.58

 rubbish

-0.58

 trade

-0.57

mage

-0.57

 polymorph

-0.56

POSITIVE LOGITS

liest

0.83

itol

0.78

interstitial

0.76

ioned

0.74

ESA

0.74

initialized

0.73

urated

0.72

acebook

0.67

oidal

0.66

ired

0.66

Activations Density 0.041%

locations or points of reference within a text

phrases indicating locations or contexts

No Comments

No Known Activations