INDEX

Explanations

references to winning trophies or awards

oai_token-act-pair · gpt-3.5-turbo

words related to trophies and awards

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

utenberg

-0.76

trak

-0.72

erman

-0.67

iquid

-0.66

ugen

-0.66

 sketch

-0.65

early

-0.64

ilipp

-0.63

ael

-0.63

ths

-0.63

POSITIVE LOGITS

 trophies

1.14

 trophy

1.11

 Trophy

0.95

ilage

0.79

 Rica

0.75

phies

0.75

OGR

0.73

 accol

0.72

brates

0.70

Congratulations

0.69

Activations Density 0.012%

references to winning trophies or awards

words related to trophies and awards

No Comments

No Known Activations