INDEX

Explanations

the word "conventional" followed by further descriptions or contexts

oai_token-act-pair · gpt-3.5-turbo

references to conventional wisdom or established norms

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

gur

-0.79

oran

-0.77

oning

-0.77

hov

-0.74

olulu

-0.73

hop

-0.71

haw

-0.71

ander

-0.70

 Wanted

-0.70

 Janeiro

-0.70

POSITIVE LOGITS

 wisdom

1.18

izations

0.85

 conventional

0.83

ties

0.83

ization

0.83

ventional

0.79

idad

0.78

arily

0.74

ities

0.74

ised

0.74

Activations Density 0.016%

the word "conventional" followed by further descriptions or contexts

references to conventional wisdom or established norms

No Comments

No Known Activations