INDEX

Explanations

Initials "Et" at the beginning of a word, typically seen in names or abbreviations

oai_token-act-pair · gpt-3.5-turbo

the word "Et" and its variations, indicating a focus on conjunctions or linking phrases in a narrative

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

biased

-0.66

anmar

-0.65

æĪ¦

-0.63

interrupted

-0.63

EStreamFrame

-0.59

¥ŀ

-0.58

against

-0.58

 narrowly

-0.57

 hawk

-0.57

esses

-0.57

POSITIVE LOGITS

iquette

1.54

ihad

1.22

udes

1.11

iology

1.03

ienne

1.03

ween

0.91

ude

0.91

ree

0.88

obic

0.87

ablish

0.86

Activations Density 0.019%

Initials "Et" at the beginning of a word, typically seen in names or abbreviations

the word "Et" and its variations, indicating a focus on conjunctions or linking phrases in a narrative

No Comments

No Known Activations