INDEX

Explanations

instances of the term "caps" along with a numeric value that might signify emphasis or intensity

oai_token-act-pair · gpt-3.5-turbo

words related to "caps" or "cap situations," indicating contexts involving limits, coverings, or championship titles

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Course

-0.79

 Answer

-0.72

Bird

-0.71

Ba

-0.70

 Dance

-0.68

 Hussain

-0.66

iate

-0.66

OU

-0.65

iri

-0.65

dden

-0.65

POSITIVE LOGITS

 caps

4.44

caps

2.67

cap

2.36

 Caps

2.00

 capped

1.93

cap

1.77

CAP

1.44

Cap

1.33

CAP

1.28

 hats

1.18

Activations Density 0.011%

instances of the term "caps" along with a numeric value that might signify emphasis or intensity

words related to "caps" or "cap situations," indicating contexts involving limits, coverings, or championship titles

No Comments

No Known Activations