INDEX

Explanations

contractions of words with 'do not'

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "don't."

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

 afore

-0.72

 VERS

-0.67

 Dise

-0.66

EStreamFrame

-0.65

pha

-0.61

DEP

-0.61

 Amph

-0.61

 Remastered

-0.60

ULT

-0.60

 Volume

-0.59

POSITIVE LOGITS

't

1.45

ned

1.14

atives

1.02

ates

0.98

uts

0.97

ning

0.96

nie

0.88

nell

0.85

kie

0.85

atters

0.84

Activations Density 0.103%

contractions of words with 'do not'

phrases that include the word "don't."

No Comments

No Known Activations