INDEX

Explanations

images or descriptions of people in different contexts or situations

oai_token-act-pair · gpt-3.5-turbo

references to images or photographs in context

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

mental

-0.82

omsky

-0.80

alities

-0.80

alties

-0.77

ances

-0.76

lan

-0.75

ulate

-0.75

iencies

-0.74

sense

-0.72

ulates

-0.72

POSITIVE LOGITS

 Thumbnails

1.06

 pictured

0.88

 IMAGES

0.83

 Scenes

0.79

 Pict

0.79

 pict

0.78

 gallery

0.76

 flanked

0.74

 circa

0.74

Pic

0.74

Activations Density 0.036%

images or descriptions of people in different contexts or situations

references to images or photographs in context

No Comments

No Known Activations