INDEX

Explanations

words related to images or designs containing images

oai_token-act-pair · gpt-3.5-turbo

references to images and image-based designs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

hurst

-0.91

cffff

-0.86

aptic

-0.86

laws

-0.81

woods

-0.79

wood

-0.74

dule

-0.73

iability

-0.73

ategory

-0.73

uckle

-0.72

POSITIVE LOGITS

 galleries

0.98

 gallery

0.97

gallery

0.91

 caption

0.90

 depicting

0.90

img

0.89

 Gallery

0.86

 images

0.85

 image

0.83

 photographed

0.82

Activations Density 0.053%

words related to images or designs containing images

references to images and image-based designs

No Comments

No Known Activations