INDEX

Explanations

official titles or designations within a context

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "official"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ĸļ

-0.98

ï¸

-0.86

aughs

-0.69

 Hath

-0.68

avery

-0.68

esville

-0.67

aws

-0.67

arcity

-0.67

ourke

-0.67

Range

-0.66

POSITIVE LOGITS

 official

1.04

official

0.88

 sanctioned

0.83

wcs

0.80

Official

0.79

 fiat

0.78

dom

0.77

 unofficial

0.77

 Official

0.75

 officially

0.75

Activations Density 0.017%

official titles or designations within a context

occurrences of the word "official"

No Comments

No Known Activations