INDEX

Explanations

tech-related terms and jargon

oai_token-act-pair · gpt-3.5-turbo

the letter 'K' and its variations within text

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

 expires

-0.70

terday

-0.69

 diapers

-0.66

 decay

-0.65

ãĥ¼ãĥĨãĤ£

-0.64

 bottleneck

-0.63

 stems

-0.63

 constitu

-0.62

 constructive

-0.62

 medieval

-0.61

POSITIVE LOGITS

EEP

1.02

won

1.01

irk

1.01

orea

0.99

essel

0.97

laus

0.97

rieg

0.97

ratom

0.96

ernel

0.95

yon

0.94

Activations Density 0.028%

tech-related terms and jargon

the letter 'K' and its variations within text

No Comments

No Known Activations