INDEX

Explanations

the name "Marcus"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Marcus."

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

boarding

-0.94

ships

-0.76

ship

-0.72

sworth

-0.71

BOOK

-0.70

eer

-0.69

 stroke

-0.69

stroke

-0.68

eared

-0.68

board

-0.68

POSITIVE LOGITS

 Aure

1.20

Marcus

0.96

ias

0.86

 Curtis

0.76

 Mari

0.75

 Marcus

0.75

Fen

0.75

 Malfoy

0.74

 Dixon

0.73

ius

0.73

Activations Density 0.015%

the name "Marcus"

mentions of the name "Marcus."

No Comments

No Known Activations