INDEX

Explanations

phrases indicating a pause or interruption in activity

oai_token-act-pair · gpt-3.5-turbo

references to breaks or pauses in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

itatively

-0.82

ifice

-0.73

omorphic

-0.73

ãĥĺãĥ©

-0.72

metics

-0.66

ificate

-0.65

ulp

-0.65

 metic

-0.63

ieval

-0.63

minist

-0.62

POSITIVE LOGITS

away

1.16

downs

1.01

neck

1.01

aways

0.94

breaks

0.91

down

0.89

break

0.88

breakers

0.87

through

0.84

waters

0.83

Activations Density 0.022%

phrases indicating a pause or interruption in activity

references to breaks or pauses in various contexts

No Comments

No Known Activations