INDEX

Explanations

the word "Dom" or variations of it

oai_token-act-pair · gpt-3.5-turbo

terms related to dominance in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

packing

-0.75

SEE

-0.66

 seeker

-0.61

 arrow

-0.59

UGH

-0.59

arson

-0.58

 arrows

-0.57

forms

-0.57

eye

-0.57

 eyes

-0.57

POSITIVE LOGITS

estic

1.40

ino

1.31

inant

1.27

inating

1.20

ination

1.18

ains

1.15

aine

1.05

inate

1.04

inated

1.00

inion

0.98

Activations Density 0.029%

the word "Dom" or variations of it

terms related to dominance in various contexts

No Comments

No Known Activations