INDEX

Explanations

mentions of the character "James Bond"

oai_token-act-pair · gpt-3.5-turbo

references to the James Bond franchise

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

OURCE

-0.88

ãĥ³ãĤ¸

-0.85

itably

-0.79

usra

-0.79

oaded

-0.78

NPR

-0.78

pace

-0.76

resents

-0.75

GBT

-0.75

ãĥ´ãĤ¡

-0.74

POSITIVE LOGITS

 Bond

1.02

 Bonds

0.83

minster

0.81

sov

0.80

eson

0.78

enegger

0.74

Rib

0.69

wick

0.69

ing

0.68

 bonds

0.68

Activations Density 0.012%

mentions of the character "James Bond"

references to the James Bond franchise

No Comments

No Known Activations