INDEX

Explanations

phrases involving comparing different things

oai_token-act-pair · gpt-3.5-turbo

instances of the word "with" indicating comparisons or relationships

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

 alone

-0.68

zin

-0.66

icious

-0.65

press

-0.65

sylvania

-0.65

entimes

-0.65

hops

-0.63

fighter

-0.63

BUG

-0.63

chief

-0.62

POSITIVE LOGITS

 regards

1.14

 regard

1.09

stood

0.97

drawn

0.94

 impunity

0.87

draw

0.86

 respect

0.86

holding

0.71

holders

0.68

trl

0.66

Activations Density 0.132%

phrases involving comparing different things

instances of the word "with" indicating comparisons or relationships

No Comments

No Known Activations