INDEX

Explanations

references to stars and related properties, especially in the context of media or entertainment

oai_token-act-pair · gpt-3.5-turbo

terms related to movies and entertainment

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

ord

-0.87

lahoma

-0.83

avier

-0.78

ucks

-0.77

opol

-0.73

roman

-0.72

onomic

-0.72

ffen

-0.71

heimer

-0.69

illance

-0.69

POSITIVE LOGITS

NING

1.34

NESS

1.31

LIN

1.30

LINE

1.30

MAN

1.27

STON

1.26

ING

1.26

TER

1.25

INESS

1.25

MENT

1.24

Activations Density 0.103%

references to stars and related properties, especially in the context of media or entertainment

terms related to movies and entertainment

No Comments

No Known Activations