INDEX

Explanations

descriptions related to visual imagery

oai_token-act-pair · gpt-3.5-turbo

references to visual elements or artworks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

vol

-0.90

bern

-0.86

ber

-0.82

clud

-0.78

cop

-0.75

TAIN

-0.72

å£«

-0.71

ney

-0.70

rament

-0.69

berman

-0.68

POSITIVE LOGITS

 imagery

0.97

 artwork

0.87

 visuals

0.86

allery

0.74

 rendition

0.73

 signage

0.72

icz

0.71

 grounding

0.70

ulkan

0.69

fulness

0.69

Activations Density 0.043%

descriptions related to visual imagery

references to visual elements or artworks

No Comments

No Known Activations