INDEX

Explanations

phrases related to being well informed or knowledgeable

oai_token-act-pair · gpt-3.5-turbo

instances of the word "informed."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ingu

-0.73

cial

-0.69

interstitial

-0.66

fighting

-0.66

 Bees

-0.65

amiya

-0.64

ayed

-0.64

amen

-0.60

Fin

-0.59

urdue

-0.59

POSITIVE LOGITS

 informed

0.83

 Consent

0.79

£ı

0.78

 antioxid

0.78

 consent

0.78

ĵĺ

0.77

 beforehand

0.76

uthor

0.73

 inform

0.73

 informs

0.72

Activations Density 0.036%

phrases related to being well informed or knowledgeable

instances of the word "informed."

No Comments

No Known Activations