INDEX

Explanations

words related to acts of power or influence

oai_token-act-pair · gpt-3.5-turbo

the "ast" suffix in various words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

yip

-0.87

pared

-0.82

perty

-0.79

BuyableInstoreAndOnline

-0.78

senal

-0.78

cci

-0.76

pton

-0.75

alties

-0.74

staking

-0.73

 tremend

-0.70

POSITIVE LOGITS

rophe

1.00

roph

0.92

ruct

0.91

efully

0.91

rophic

0.90

rop

0.86

ro

0.83

roid

0.83

odon

0.83

rians

0.83

Activations Density 0.013%

words related to acts of power or influence

the "ast" suffix in various words

No Comments

No Known Activations