INDEX

Explanations

proper nouns, especially names, that contain the word "ord"

oai_token-act-pair · gpt-3.5-turbo

variations of the word "word" and related terms in different contexts

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

itÃ©

-0.71

cling

-0.67

ted

-0.66

fox

-0.65

Fu

-0.64

gets

-0.64

mitt

-0.64

 grease

-0.64

apo

-0.63

tty

-0.62

POSITIVE LOGITS

inarily

1.06

ORD

1.02

nance

0.98

igham

0.91

inances

0.90

itudinal

0.84

LESS

0.83

LEASE

0.83

inator

0.83

 ORDER

0.78

Activations Density 0.006%

proper nouns, especially names, that contain the word "ord"

variations of the word "word" and related terms in different contexts

No Comments

No Known Activations