INDEX

Explanations

phrases related to services or benefits

oai_token-act-pair · gpt-3.5-turbo

instances of high numeric values associated with various services or features

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

vati

-0.59

uph

-0.58

etheless

-0.57

ertodd

-0.57

yt

-0.56

 renegoti

-0.55

 citiz

-0.55

 tradem

-0.55

 helicop

-0.55

lihood

-0.54

POSITIVE LOGITS

Contents

0.91

SPONSORED

0.86

Details

0.78

Origin

0.78

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

0.76

Specifically

0.76

DEN

0.76

ADVERTISEMENT

0.76

Interested

0.75

Guest

0.72

Activations Density 0.336%

phrases related to services or benefits

instances of high numeric values associated with various services or features

No Comments

No Known Activations