INDEX

Explanations

words related to a specific person's name or title 'Ferrer'

oai_token-act-pair · gpt-3.5-turbo

terms related to individuals in a professional or academic context

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 lung

-0.64

 adject

-0.62

 moons

-0.58

ang

-0.58

 mistakes

-0.56

bay

-0.56

MO

-0.54

 flash

-0.54

 feathers

-0.54

 encaps

-0.53

POSITIVE LOGITS

rer

4.95

rers

3.87

erer

1.37

ring

1.25

reth

1.24

ller

1.22

rera

1.21

arer

1.15

rals

1.14

rero

1.10

Activations Density 0.017%

words related to a specific person's name or title 'Ferrer'

terms related to individuals in a professional or academic context

No Comments

No Known Activations