INDEX

Explanations

entertainment-related words and mentions of specific venues or performers

oai_token-act-pair · gpt-3.5-turbo

references to films and movies

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

rique

-0.75

owder

-0.69

riel

-0.68

ntil

-0.64

 transitions

-0.63

lished

-0.62

 workflow

-0.61

 Helsinki

-0.60

 transitioning

-0.60

aining

-0.59

POSITIVE LOGITS

lihood

0.84

itarian

0.82

enegger

0.79

culosis

0.76

rities

0.70

 billboards

0.69

 Festival

0.69

 ridic

0.68

goers

0.64

 Jeep

0.64

Activations Density 0.237%

entertainment-related words and mentions of specific venues or performers

references to films and movies

No Comments

No Known Activations