INDEX

Explanations

photographs or images referenced within text

oai_token-act-pair · gpt-3.5-turbo

references to images or pictures in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gew

-0.89

indal

-0.77

oing

-0.77

âĵĺ

-0.73

iosity

-0.72

aciously

-0.72

hip

-0.72

bard

-0.71

itability

-0.68

vernment

-0.68

POSITIVE LOGITS

lihood

0.99

Picture

0.78

 Picture

0.75

 IMAGES

0.75

picture

0.73

pict

0.70

Pict

0.69

 Noir

0.68

 Siege

0.68

 Animation

0.65

Activations Density 0.011%

photographs or images referenced within text

references to images or pictures in the text

No Comments

No Known Activations