INDEX

Explanations

mentions of the word "Real."

oai_token-act-pair · gpt-3.5-turbo

references to the term "Real" and its associated contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

du

-0.73

 mustard

-0.70

Pes

-0.68

 joint

-0.66

-0.63

 concussion

-0.63

 Nicola

-0.62

BP

-0.62

 Aden

-0.62

POSITIVE LOGITS

Real

3.30

Way

1.24

Force

1.23

True

1.22

real

1.18

Hour

1.18

River

1.16

Filter

1.15

Rap

1.15

Owner

1.14

Activations Density 0.031%

mentions of the word "Real."

references to the term "Real" and its associated contexts

No Comments

No Known Activations