INDEX

Explanations

instances of the word "full"

oai_token-act-pair · gpt-3.5-turbo

phrases indicating totality or completeness

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

apters

-0.73

 Disciple

-0.72

 Shack

-0.69

 Royale

-0.67

 Pere

-0.67

eers

-0.66

 Petty

-0.66

Hop

-0.65

pher

-0.65

 Murd

-0.64

POSITIVE LOGITS

erton

1.01

screen

0.90

 extent

0.90

 brunt

0.88

 complement

0.83

heartedly

0.83

 employment

0.83

 throttle

0.81

empt

0.80

 fled

0.80

Activations Density 0.030%

instances of the word "full"

phrases indicating totality or completeness

No Comments

No Known Activations