INDEX

Explanations

the prefix "ex-" with different following words

oai_token-act-pair · gpt-3.5-turbo

prefixes related to the concept of 'ex' or former status

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

trop

-0.90

abbage

-0.87

halla

-0.75

uckland

-0.74

nen

-0.72

trak

-0.71

jong

-0.71

kefeller

-0.70

velt

-0.70

zai

-0.68

POSITIVE LOGITS

ex

3.51

Ex

1.71

ex

1.66

Ex

1.36

EX

1.28

 former

1.28

exp

1.26

 estranged

1.24

exh

1.19

EX

1.15

Activations Density 0.017%

the prefix "ex-" with different following words

prefixes related to the concept of 'ex' or former status

No Comments

No Known Activations