INDEX

Explanations

experiences shared by individuals

oai_token-act-pair · gpt-3.5-turbo

instances of the word "experiences."

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

 tumor

-0.69

Cou

-0.65

sub

-0.65

vous

-0.64

hatt

-0.62

egal

-0.61

 withholding

-0.61

 spare

-0.60

cut

-0.59

 clot

-0.59

POSITIVE LOGITS

 experiences

1.34

iences

1.12

 experien

1.09

 Experience

0.95

ivities

0.89

 experience

0.89

 Exper

0.89

Experience

0.84

olutions

0.82

afety

0.78

Activations Density 0.006%

experiences shared by individuals

instances of the word "experiences."

No Comments

No Known Activations