INDEX

Explanations

references to cinemas or movies

oai_token-act-pair · gpt-3.5-turbo

references to cinemas and their various aspects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

lies

-0.78

lynn

-0.69

essee

-0.67

lying

-0.66

ually

-0.65

ional

-0.65

wid

-0.64

hips

-0.64

iang

-0.64

bands

-0.64

POSITIVE LOGITS

ovies

0.94

 cinem

0.94

goers

0.89

atography

0.88

 cinema

0.88

 marqu

0.83

 theaters

0.82

 premie

0.79

 theat

0.79

isoft

0.77

Activations Density 0.026%

references to cinemas or movies

references to cinemas and their various aspects

No Comments

No Known Activations